The amyloid cascade hypothesis, the most recognized hypothesis to date, was proposed by British scientists Hardy and Higgins (1992). The β-amyloid precursor protein (APP) is converted into the polypeptide Aβ, which has 39–43 residues, by β-secretase and γ-secretase. Aβ exists in the brain in multiple forms, including soluble monomers, oligomers, protofibrils, and insoluble Aβ fibers in amyloid plaques (Hampel et al., 2021). Additionally, amyloid oligomers are soluble and spread throughout the brain. The two Aβ isoforms that contain 40 residues each, Aβ40 and Aβ42 are thought to be the most significant among all isoforms. The two additional residues at the C-terminus of Aβ42 are the only difference in amino acid residues between these two isoforms. However, they differ in metabolism, aggregation mechanisms, and toxicities. Particularly, Aβ40 is more likely to aggregate and form oligomers, with strong neurotoxicity, whereas Aβ40 has higher solubility and easier to remove cells. Moreover, the abundance of Aβ42 is the highest in amyloid plaques of brain samples of patients with AD (Lesné et al., 2006).

The Aβ plaque hypothesis, which suggests that Aβ protein accumulation in the brain is a key factor in AD progression, is the foundation for many drugs developed today. One of the most promising AD treatments includes immunotherapies focused on Aβ, which include strategies to decrease Aβ production, increase Aβ clearance, and inhibit Aβ aggregation. A novel fusion protein called αAβ-Gas6, developed by Jung and his team, can successfully promote Aβ clearance because it works through a different mechanism than Aβ-based mAbs (Jung et al., 2022). In addition, αAβ-Gas6 eliminated Aβ plaques in AD models without causing the neurotoxic inflammatory side effects associated with a conventional antibody treatment.

However, there are numerous negative trials of Aβ-targeting drugs for mild to moderate AD over the past 20 years, and the limited efficacy and no benefit also was observed in many trials, which have raised the question of whether Aβ is the right target (Rafii and Aisen, 2023). Recently, Aβ-targeting drugs has made substantial progress. A systematic review shows that aducanumab reduced brain Aβ plaques in a time- and dose-dependent manner, and aducanumab reduced the severity of clinical dementia in the high-dose treated group (Rahman et al., 2023). In addition, lecanemab was found to produce marked lowering of Aβ plaques based on evidence derived from amyloid positron emission tomography, and lecanemab shows “clinical benefit” for the AD treatment, paving the way for the medication to be considered for full FDA approval (Mahase, 2023; van Dyck et al., 2023).

Along with Aβ and cerebrovascular amyloid angiopathy, neurofibrillary tangles (NFTs) and neuropil threads are neuropathological confirmations in diagnosing AD. The aberrant intraneuronal Tau protein, which is highly neurotoxic and impairs cognition, makes up the entire structure of NFTs. An overview of the discovery of Tau proteins dates back to 1974 when American scientist Iqbal et al. isolated NFT and paired helical filament (PHF) proteins from the brains of patients with AD for the first time (Iqbal et al., 1974). In the following year, Weingarten et al. (1975) isolated Tau protein from pig brains. Then, Grundke-Iqbal et al. (1986) found that PHFs consist of Tau protein, which is significantly more phosphorylated than normal human brain Tau. Tau protein is a microtubule-associated protein that is highly soluble and abundant in the neurons of the CNS. It primarily affects the distal end of axons, preserving the stability and flexibility of microtubules while also regulating axonal transport and promoting actin filament formation (Venkatramani and Panda, 2019). Normal Tau carries multiple phosphate groups in its microtubule assembly domain, but phosphorylation of tau reduces its affinity for microtubules and begins to aggregate in its hyperphosphorylated form (Mohammadi et al., 2023). Elevated phosphorylation and aggregation of Tau are widely considered pathological hallmarks in AD (Wegmann et al., 2021).

Dr. Mielke found that peripheral blood phosphorylated Tau at threonine-181 (P-Tau 181) and phosphorylated Tau at threonine-217 (P-Tau 217) levels are reliable indicators of cerebral amyloid lesions and can be used as blood-based biomarkers to distinguish AD from other neurodegenerative diseases (Mielke et al., 2022). Stabilizing the structure of tubulin, inhibiting the phosphorylation and aggregation of Tau, and promoting the degradation of aggregated Tau are the major steps in developing drugs that target Tau. Fang et al. found that sildenafil has neuroprotective and neurorestorative roles by promoting neurite growth and reducing aberrant Tau protein phosphorylation (Fang et al., 2021).

Neuroinflammation is another distinguishing hallmark of the brain in AD patients, along with Aβ plaques and NFTs, and it is also thought to play a role in the progression of AD (Si et al., 2023). Study has shown that the serum levels of inflammatory cytokines, such as interleukin-6 (IL-6), IL-1β, tumor necrosis factor α (TNF-α), and interferon-γ, are significantly higher in patients with AD than in healthy individuals (Demirci et al., 2017). The risk of AD is increased by chronic inflammation in the brain, which is mediated by interactions between the immune system and the CNS. Before Aβ plaque deposition, neuroinflammation occurs in the very early stages of AD. Recent studies suggest that as the disease advances, neuroinflammation promotes the synthesis of new Aβ, and increased Aβ levels in the brain can induce and exacerbate neuroinflammation (Bronzuoli et al., 2016; Yang, 2019). This leads to a vicious cycle between Aβ accumulation and neuroinflammation.

The resident brain macrophages known as microglia, detected in certain brain areas in patients with AD, have been found to promote phagocytosis to monitor and protect neural tissue. Hexokinase 2 (HK2), a glycolytic enzyme, is shown to be highly upregulated in microglia in a mouse model of AD (Leng et al., 2022). The increased activity of HK2 resulted in insufficient energy production in these cells, decreasing microglial phagocytosis. However, HK2 inhibition causes a switch from glucose to lipid metabolism to ensure efficient energy, leading to microglia with strong phagocytic activity to clear Aβ plaques and improve cognitive function (Choi and Mook-Jung, 2022; Leng et al., 2022). However, there is a debate concerning the role of microglia in AD, and when these cells lose their ability to maintain homeostasis, they may have harmful effects. The pyrin-domain containing 3 (NLRP3) inflammasome is one of the multiprotein complexes that microglia can produce. When NLRP3 is activated, it can cleave the downstream inflammatory factor IL-1β from the precursor to the mature state and release it into the extracellular space, which causes neuroinflammation (Heneka et al., 2018). Further studies have found that Aβ-induced Tau lesions depend on the NLRP3 inflammasome, and the NLRP3 inflammasome plays a role in the development of Tau lesions and AD. IL-1β is also highly expressed in the cortex of patients with frontotemporal dementia, suggesting inflammasome activation.

Following the interferon-induced transmembrane protein 3 (IFITM3) in aging mice compared to healthy mice, another study found that IFITM3 is upregulated in tissue samples from patients with advanced AD (Hur et al., 2020). Anti-inflammatory drug administration that can lower IFITM3 levels also reduces the production of Aβ, delaying the onset and progression of AD. IL-3 is a crucial mediator of astrocyte-microglia crosstalk and a node of AD therapeutic intervention, and the study of Harvard University discovered that it had a protective effect in an AD mouse model (McAlpine et al., 2021). These findings suggest that regulating neuroinflammation might be a novel treatment option for AD.

In the brains of patients with AD, there is a general decrease in the levels of neurotransmitters, particularly those involved in learning and memory, such as the acetylcholine (Ach) system, amino acids, monoamine system, neuropeptides, and other neurotransmitters (such as γ-aminobutyric acid, glutamate, epinephrine, etc.). Previous study has confirmed that Ach and choline acetyltransferase levels in the hippocampus and cortex of patients with AD are significantly decreased (DeKosky et al., 2002). Current standard treatments for patients with mild-to-moderate AD include cholinesterase inhibitors such as donepezil, rivarastine, galantamine, and huperzine A. The basic mechanism is to increase the concentration of Ach in the brain and prevent cholinesterase (Snowden et al., 2019). Studies have shown that elderly patients with cholinesterase inhibitors can also improve cognitive function and promote the release of dopamine, norepinephrine, epinephrine, and other neurotransmitters in the brain (Wynn and Cummings, 2004). However, cholinesterase inhibitors can cause side effects such as nausea, vomiting, and diarrhea (Hogan, 2014).

There is growing evidence that oxidative stress plays a key role in the pathogenesis of AD (Sultana and Butterfield, 2010). Reactive oxygen species (ROS) can cause the oxidation of intracellular biological macromolecules (such as nucleic acids, membrane lipids, and cellular proteins), which can negatively impact synaptic activity, damage nerve cells, and result in cognitive dysfunction. Loss of mitochondrial function, altered metal homeostasis, and decreased antioxidant capacity can increase ROS production. ROS also promotes aberrant Aβ deposition and Tau hyperphosphorylation, which aggravates mitochondrial dysfunction and ROS production, resulting in AD and creating a vicious cycle (Kent et al., 2020; Wang et al., 2020). Therefore, inhibiting mitochondrial ROS overproduction, metal ion chelation therapy, and increasing antioxidant capacity can improve the symptoms of AD.

Mitophagy is an autophagy mechanism that selectively eliminates defective mitochondria, preserving cellular homeostasis and energy metabolism in cells. Early cellular homeostasis in AD involves mitochondrial damage, and cell ROS levels result in abnormal mitophagy and autophagy. A recent study has shown a previously unreported pattern of autophagic stress, known as poisonous anthos flower, associated with impaired neuron autophagy and present in AD brains (Lee et al., 2022). Although the molecular basis for mitophagy and autophagy dysfunction in AD is unknown, this pattern of autophagic stress has been linked to AD (Lee and Nixon, 2022). Additionally, using imaging and histochemical methods, they showed that the neurons are the source of the great majority of senile plaques in AD mice models, leading to a re-evaluation of the generally understood sequence of events in plaque formation in AD (Lee et al., 2022).

Recent studies show that the most common genetic risk factor for AD is the lipid and cholesterol gene apolipoprotein E4 (APOE4) (Simon et al., 2012; Ryu et al., 2019; Safieh et al., 2019). One copy of the APOE4 allele increases the risk of getting AD by three to four times, while two copies can increase the risk by 8–12 times (Simon et al., 2012). According to the studies, the APOE4 allele is present in 50%–60% of patients with AD (Abushakra et al., 2016; Ryu et al., 2019). APOE4 drastically changed the signaling pathways involved in cholesterol homeostasis and transport, hampered myelination, and impacted cognitive function in mice. In mice with APOE4, pharmacological intervention to lessen this effect enhanced learning and memory. This study provides a prospective treatment strategy for AD by demonstrating a functional relationship between APOE4, cholesterol, myelination, and memory (Blanchard et al., 2022). In addition, the APOE4 gene causes glia-specific cellular and cell-free dysregulation, which raises the risk for AD (Tcw et al., 2022). The studies mentioned above present a new strategy of treating AD.

In a promising nanoscale drug delivery system, polymer NPs can encapsulate drugs within a biodegradable polymeric matrix or alter their surface to provide a controlled release and improved bioavailability. Polymer NPs can range from 10 to 1,000 nm. Poly (lactic-co-glycolic acid) (PLGA), polyethylene glycol (PEG), poly-L-lysine, polylactic acid (PLA), polyethyleneimine, poly (acrylate-co-alkyl acrylate) (PACA), poly (butyl cyanoacrylate), polycaprolactone, chitosan, and gelatin are commonly used to make polymer NPs (Zhang W. et al., 2021). The basic idea behind this class of nanoscale drug delivery systems is to encapsulate drugs within the matrix of the carrier material or to modify them by covalent linkage/adsorption on the surface of the carrier material. The excellent biocompatibility, high bioavailability, and controllable biodegradability of polymer NPs span a broad spectrum. The technology is currently one of the most active research areas, and has enormous potential for application in nanomedicine.

Extensive studies shows that curcumin has anti-Aβ formation, aggregation, anti-inflammatory, and antioxidant effects that can help middle-aged and older people with mild cognitive impairment improve their memory (Fan et al., 2018; Yang et al., 2018). Therefore, Fan et al. designed a brain-targeted curcumin NPs system containing PLGA-PEG combined with B6 peptide and curcumin. The results showed that compared with curcumin administration, curcumin NPs could reduce hippocampal Aβ deposition and excessive Tau protein phosphorylation, significantly improving the spatial learning and memory ability of APP/Presenilin 1 (PS1) mice (Fan et al., 2018). Additionally, Yang et al. developed bovine serum albumin and chitosan loaded with curcumin. They found that the NPs could effectively increase the penetration of drugs through the BBB, promote microglia activation, and accelerate the phagocytosis of Aβ42 (Yang et al., 2018).

Donepezil is primarily used to treat mild and moderate AD. However, the drug’s concentration in the brain is restricted due to poor penetration. Therefore, to prepare organic NPs loaded with donepenzil, Liu et al. introduced regadenoson (Reg) at the end of PEG2000. Reg-NPs could enhance drug delivery into the brain, and activate the PKG/PI3K and NO/cGMP signaling pathways. The NPs improved learning and memory in Aβ-induced AD mice as compared with donepezil at the same dose (Liu et al., 2021).

Depending on how they charge, polymer NPs can be divided into cationic, anionic, and zwitterionic polymers. Positively charged polymers are amine-containing polymers that can electrostatically adsorb or encapsulate negatively charged nucleic acids. Wang et al. constructed PEG-polydimethylaminoethyl methacrylate (cationic polymer) NPs loaded with β-site amyloid precursor protein cleaving enzyme 1 (BACE1) to treat APP/PS1 transgenic mice. According to the findings, these NPs significantly decreased BACE1 mRNA and Aβ plaques, inhibited Tau phosphorylation, and promoted hippocampal neurogenesis (Wang et al., 2018).

Liposomes are the earliest nanomedicine delivery system to proceed from concept to preclinical research and clinical application. Liposomes mainly comprise natural lipids, usually phospholipids, ionizable, cholesterol, and pegylated. Liposomes have no pharmacological activity, low toxicity, and good biodegradability, mobility, and biocompatibility (Patel et al., 2021). Therefore, they promise vehicles to deliver therapeutics, such as antitumor agents and anti-inflammatory, antibiotic, antifungal, anesthetic, and other drugs. Moreover, liposomes’ hydrophilic and hydrophobic components can encapsulate water-soluble and lipid-soluble drugs, improving their bioavailability. Several liposomal drugs have been approved for marketing worldwide. Chen et al. modified liposomes with transferrin to improve the brain delivery of α-mangostin, a polyphenolic xanthone effective in treating AD. In addition, the transferred liposomes could successfully transport α-mangostin across the BBB in vitro and in vivo (Chen et al., 2016).

Tannic acid is a multi-branched polyphenol molecule that effectively inhibits Tau aggregation. Hu et al. coated the liposomes with Tween 80 and added tannic acid to the centre of each one, tannic acid liposomes can cross the BBB and deliver drugs effectively (Hu et al., 2020). In addition to internal drug loading, phospholipids can react with the drug to synthesize the phospholipid prodrug, which can then assemble with other lipids to form liposomes with surface modification of the drug and achieve an anti-AD effect. Introducing phosphatidic acid with Aβ affinity into the liposome component can also endow the liposome with certain AD efficacy. Mancini et al. used phosphatidic acid to create liposomes (Mancini et al., 2016). The modified ApoE-derived peptide was also added to the surface of the liposomes to increase BBB permeability. The outcomes demonstrated that the generated liposomes might delay the cognitive function decline in mice. Furthermore, the liposomes were optimized for drug delivery and treatment, after administration, the area and number of Aβ plaques in the cortex and hippocampus were reduced by 29% and 31%, respectively. The above studies have shown that liposomes help drugs enter the brain, making them the most promising drug-carrier system for treating AD.

SLNs are solid core lipid nanocarriers, which are developed to hold both hydrophilic and lipophilic drugs (Paliwal et al., 2020). The particle size ranges between 10 and 1,000 nm, and fatty acids, fatty alcohols, and phospholipids comprise most of the components in SLNs. The drug is encapsulated or trapped in the lipid core to create a solid colloidal drug delivery system. Due to their improved physical stability, higher loading capacity, better bioavailability, and ease of large-scale production, SLNs are excellent for poorly soluble drugs (e.g., adriamycin and cyclosporine). Therefore, it has been widely used to create various dosage forms, including oral, injection, transdermal, and pulmonary delivery, and has broad prospects in developing new drugs for AD.

Ferulic acid is an active component of herbal medicines, which can inhibit Aβ aggregation and improve the viability of AD cell models in vitro and is considered as a potential drug for AD treatment (Mancuso and Santangelo, 2014; Kikugawa et al., 2016). However, Ferulic acid has the characteristics of poor water solubility and limited brain permeability. Therefore, using Compritol as a lipid and Tween 80 as a surfactant, Saini et al. constructed SLNs for ferulic acid. SLNs significantly improved learning and memory functions in AD rats and had an excellent nasal mucosal adhesion and penetration ability (Saini et al., 2021). Additionally, Vedagiri and Sumathi created SLNs loaded with chrysin, and found that SLNs could increase the bioavailability of chrysin and decrease the neuronal damage caused by Aβ25-35 (Vedagiri and Sumathi, 2016). Yusuf et al. created piperine-loaded SLNs in a different study. They found that SLNs could improve Aβ plaque deposition by reducing oxidative stress and cholinergic degradation in the cortex of AD rats induced by atraconic acid (Yusuf et al., 2013). Additionally, Dhawan et al. created quercetin-loaded SLNs, they found that compared to pure quercetin, SLNs significantly decreased their antioxidant capacity and prevented the neurodegeneration induced by alumina in rats (Dhawan et al., 2011). According to the studies above, SLNs can cross the BBB and increase the bioavailability of insoluble drugs, and it have good potential as nanocarriers for the treatment of AD.

Highly branching globular polymers, known as dendritic NPs, often have compact three-dimensional structures with a central core and corona. They mostly use dendrimers made of polyamide-amine (PAMAM) (Janaszewska et al., 2018). Their unique three-dimensional structure gives them unique properties, such as surface modification of functional groups, spherical nanostructures, internal hydrophilic or hydrophobic cavities, and a low polydispersity index (Li et al., 2021). The internal core of dendrimers, which contains functional groups, provides space and binding capabilities for several drug molecules, including chemical compounds, peptides, genes, and other substances.

A novel class of nanocomposites, known as dendritic NPs, has been applied in AD drug development. The efficient passage of dendritic NPs through the BBB can be attributed to its lipophilicity (Kaurav et al., 2023). Modifying the surface of dendrimers will improve biocompatibility and enhance drug transport through the BBB (Zhu et al., 2019). Kumari et al. demonstrated how the preparation and synthesis of hyperbranched dendritic polymers, such as PACA, PLA, and PLGA, can inhibit fiber formation and have a specific therapeutic effect on AD (Kumari et al., 2010). PAMAM dendrimers can disrupt existing fibrils and interfere with the development of Aβ1-28 in AD brain, which can delay the progression of AD (Klajnert et al., 2006). Additionally, Klementieva et al. found that maltose (MAL)-modified polypropylamine (PPI) dendrimers can prevent Aβ1-40 fibrosis. They include PPI-G5-MAL, which can block the formation of amyloid fibers, and PPI-G4-MAL, which can reduce amyloid toxicity by aggregating fibers (Klementieva et al., 2011). Hence, dendritic NPs may have promising application prospects in treating AD.

The size of metal NPs is usually between 1 and 100 nm, and they typically have a functional organic molecules shell around a metal core. Presently, pure metals, including gold, silver, copper, zinc, etc., or their oxides make up metal oxide NPs (Jamkhande et al., 2019). For example, cerium and iron oxides make metal oxide NPs, while molybdenum disulfide makes metal sulfoxide NPs. Drug molecules can be coupled to the surface of metal NPs due to their dense spherical carrier structures, large surface area, and potential for drug loading. Additionally, metal NPs are used in targeted drug delivery approaches for CNS diseases because of their excellent biocompatibility, storage stability, ease of synthesis, and few side effects. However, metal NPs are difficult to degrade in vivo, and their safety hazards limit their clinical application.

Recently, developed metal-based NPs with higher stability and permeability to tissues like cancer tissues have been used as carriers for photosensitizer drugs. These carriers can be made of gold or magnetically responsive carriers. Nanomaterial formulations have also been used to investigate the targeted delivery of CNS drugs further. Cerium oxide (CeO₂) NPs exhibit antioxidant, anti-interference, and anti-apoptosis properties. In addition, CeO₂-NPs existing in the +3 and +4 valence states of cerium have free radical scavenging activity, creating an antioxidant microenvironment and promoting the repair of nerves and other cells (Das et al., 2013). The ability of CeO₂-NPs with a size <5 nm to mimic catalase and superoxide dismutase through the reversible binding of oxygen atoms and conversion between Ce³⁺ (reducing) and Ce⁴⁺ (oxidizing) states on their surface has also been demonstrated (Celardo et al., 2011).

The lipophilic cation triphenyl phosphonium can target mitochondria. Studies in the five-familial AD transgenic mouse model have shown that cerium NPs conjugated with TPP can increase mitochondrial concentrations and inhibit neuronal death (Biswas et al., 2012). Another study linked two peptides with the KLVFF sequence (residues 16–20), which corresponds to the hydrophobic core of Aβ and plays a significant role in the formation of the β-sheet structure with ultrasmall gadolinium nanoparticles (AGuIX-NPs). Furthermore, this fragment can also bind to Aβ, affecting the formation of plaques (Lowe et al., 2001).

A magnetic core (such as iron, nickel, cobalt, and their oxides) and a surfactant makeup magnetic NPs combine the benefits of magnetic materials and nanomaterials. Due to their unique superparamagnetic characteristics, magnetic NPs can be guided and deliver drugs to specific areas when exposed to a magnetic field (Materón et al., 2021; Zhang P. et al., 2023). Iron, nickel, and cobalt are frequently used as magnetic materials, iron and iron oxide are the most common materials. Due to its biocompatibility and biodegradability, iron oxide NPs (Fe₃O₄ or Fe₂O₃) have been used in biomedical research extensively. However, the homogeneity of magnetic NPs is difficult to establish throughout the large-scale synthesis process, and the variation between various production batches is high, which can have major safety problems (Vangijzegem et al., 2019).