Amyloid-Beta (Aβ) is a crucial biomarker in AD, with its accumulation being a central feature in disease progression. Initially found in the neocortex of pre-clinical AD patients, Aβ plaques spread to regions including the hippocampus and basal ganglia in later stages, and eventually to the cerebellar cortex and lower brain stem in advanced cases (Kepp et al., 2023). Some amyloid imaging studies indicate the presence of amyloid deposits in many individuals without clinical symptoms, and conversely, the absence of significant amyloid deposits in some AD patients. However, despite these variations, it is generally agreed that Aβ is a major drive of AD, and that the accumulation of Aβ amyloid fibrils can evolve into senile plaques. These plaques are implicated in neurotoxicity and the initiation of tau pathology, which in turn leads to the death of neuronal cells and neurodegeneration characteristic of AD.

The amyloid cascade hypothesis, a prominent theory in AD research, posits that the pathological production and accumulation of Aβ peptides are central to the disease’s pathogenesis. This hypothesis continues to guide much of the global research and therapeutic approaches for AD. In the context of AD, there is an aberrant increase in the cleavage of β-amyloid precursor protein (APP) by β-secretases, followed by γ-secretases, leading to the generation of Aβ peptides of differing lengths. This process involves sequential enzymatic cleavages by β- and γ-secretases. Notably, the cleavage site of another enzyme involved in APP processing, α-secretase, exists within the Aβ sequence, thereby preventing the formation of Aβ. It is hypothesized that the longer forms of Aβ, such as Aβ42, are more pathogenic due to their increased hydrophobic nature, propensity for oligomerization, and tendency to form amyloid plaques. (Tramutola et al., 2015; Tiwari et al., 2019; Kepp et al., 2023). Several studies have associated increased Aβ42 production with increased oxidative stress and autophagy impairment to produce proteostasis dysregulation in the brain (Di Domenico et al., 2013; Tramutola et al., 2015). Moreover, it may block synaptic transmission, increase neuroinflammation, trigger neuronal cell death, thereby contributing to the pathology of AD.

mTOR activation is commonly associated with AD progression and worsened amyloidogenic Aβ synthesis and deposition. Postmortem biopsies from AD patients have shown heightened activating phosphorylation of mTOR, increased phosphorylation of mTOR downstream targets (4E-BP1 and S6K) (X. Li et al., 2005; Tramutola et al., 2015), as well as hyperactivity of the phosphoinositide 3-kinase (PI3K)/Akt/mTOR phosphorylation cascade. Brain tissues diagnosed with mild cognitive impairment (MCI) and AD show higher mTOR activating phosphorylation than pre-clinical AD. This demonstrates a correlation between mTOR activity and disease severity as marked by Braak’s staging. Recent analyses on general population and AD patient GWAS data concerning (PI3K)/Akt/mTOR pathway proteins (Akt, S6K, eIF4E) suggest that genetic backgrounds with elevated mTOR-dependent response is causally linked to AD incidences (H. Y. Cai et al., 2023).

Recent studies have highlighted the role of mTOR in modulating Aβ levels through the regulation of autophagy. This is substantiated by postmortem biopsies, which reveal diminished autophagy in AD and MCI patients. This reduction is linked to mTOR-mediated inhibition of key autophagy regulators, such as unc-51-like kinase 1 (ULK1) and transcription factor EB (TFEB) (H. Y. Cai et al., 2023). ULK1 is essential to autophagy initiation and autophagosome formation, and TFEB is a master regulator of lysosomal biogenesis. Autophagosomes engulf cellular components targeted for removal, then fuse with lysosomes for hydrolytic degradation. Quantitative evidence shows a decrease of at least 25% in autophagy markers LC-3 and Beclin one in these patients, indicating significant autophagy impairments. Corroborating this, previous studies have identified a strong correlation between elevated concentrations of Aβ, particularly Aβ42, and pronounced reductions in autophagy across all clinical groups (Tiwari et al., 2019). Suelves et al. further support this association by demonstrating intraneuronal Aβ accumulation as a consequence of autophagic defects in both AD patients and the 5xFAD AD mouse model (Suelves et al., 2023). Complementing these findings, Caccamo et al. observed increased S6K phosphorylation in the 7PA2 immortalized cell line transfected with mutant amyloid precursor protein (APP), and in the brains of 3xTg-AD model mice. Likewise, this mTOR activation exerted pro-disease progression effects via autophagy inhibition (Caccamo et al., 2010).

Interestingly, genetically preventing Aβ accumulation effectively reduced mTOR activity (Caccamo et al., 2011) and upregulated autophagy. Primary cortical neurons treated with Aβ oligomers corroborate these results and indicate that Aβ can increase levels of p- Akt^S473 and phospho-mTOR^s2448 (Bhaskar et al., 2009), suggesting a cascade-like increase in Aβ production. Moreover, when naturally secreted Aβ was injected into the hippocampus of wild-type mice, it is capable of enhancing mTOR signalling in their brains via proline-rich Akt substrate 40 (PRAS40) (Caccamo et al., 2011). This finding underscores the intricate relationship between Aβ and mTOR activation (Tiwari et al., 2019; Kepp et al., 2023) (Extensively Reviewed in (Z. Cai et al., 2015; Mueed et al., 2018)) to pose a self-perpetuating Aβ production cycles in Alzheimer’s disease pathology.

However, it is worth noting that a few studies have reported contrasting findings. Shi and colleagues observed boosted Aβ clearance in mice of both sexes, after mTOR activation caused by the selective loss of microglial Tsc1, a negative regulator of mTOR (Shi et al., 2022). These physiological changes were accompanied by an improvement in cognitive functions in the 5XFAD AD mouse model, suggesting a potential negative correlation between mTOR activation and Aβ plaque burden level (Shi et al., 2022).

Historical evidence in mice models largely shows that reducing mTOR activity via traditional mTOR inhibitors including rapamycin and rapalogs restores Aβ pathogenesis and cognitive capacities, as summarized in Table 1 (Caccamo et al., 2010; Cassano et al., 2019; Jiang et al., 2014a; Singh et al., 2017b; Van Skike et al., 2021). Spilman et al. first showed that long-term inhibition of mTOR pathway through rapamycin administration (oral, 2.24 mg/kg/day, 13 weeks) lowered Aβ42 levels in transgenic PDAPP(J20) mice, an AD model that overexpresses mutant human APP V717F (Spilman et al., 2010). Furthermore, rapamycin treatment is shown to improve learning and restore spatial memory via the Morris Water Maze (MWM), suggesting that mTOR inhibition blocked or delayed Alzheimer’s symptoms onset in the PDAPP(J20) model. Importantly, these effects were accompanied by autophagy upregulation. Administration of the autophagy inhibitor 3-methyladenine removed the positive effects of rapamycin on Aβ42 levels, in both 3xTg-AD and 7PA2 cells. The outcomes were reciprocated with similar experimental protocols on the Apolipoprotein E ɛ4 (APOE4) transgenic mice as a late-onset AD model (Lin et al., 2017). Recent evidence show that rapamycin not only reduces Aβ level, but also abrogates other pathologies induced by Aβ deposition. Van Skike et al. demonstrated in their experiment on PDAPP(J20) that cerebral blood flow, blood-brain barrier breakdown, and neurovascular uncoupling were partly mitigated via the mTOR-Aβ axis (Van Skike et al., 2018; Van Skike et al., 2021), supporting previous findings (Lin et al., 2017).

Cell culture studies substantiate the role of mTOR inhibitors in mitigating Aβ pathogenesis. Jiang et al. reported that temsirolimus facilitated the activation of autophagy in cells that overexpressed human APP (HEK293-APP695). This activation was evident from changes in biomarkers, with P62 reduced by 55% and microtubule-associated protein one light chain 3 (LC3)-II increased 2.1-fold, respectively. This autophagic activation is responsible for reducing the Aβ42 levels; Aβ42 levels remained unchanged when Atg5, an essential component of the autophagic pathway, was knocked down (T. Jiang et al., 2014a). Singh et al. reported similar results in the Aβ-intoxicated SH-SY5Y neuronal cell line, demonstrating that mTOR inhibition using rapamycin alleviates Aβ-induced oxidative stress, apoptosis, and neurotoxicity in an autophagy-dependent manner (Singh et al., 2017a).

However, few reports had reached inconsistent conclusions (Carosi and Sargeant, 2023; Gourmaud et al., 2022; J. H; Lee et al., 2022; Shi et al., 2022) which calls for caution in clinical translation. Zhang et al. describe rapamycin treatment in APP-transfected murine neuronal like cells (N2a) and Tg2576 mice models to increase Aβ accumulation (Zhang et al., 2010). Other studies on the five times familial AD (5XFAD) mouse model show that although rapamycin may attenuate memory deficits in five times familial AD mouse model (5XFAD) with seizure-induced exacerbation, treatment did not alter, or even increased, hippocampal Aβ42 levels (Singh et al., 2017a; Shi et al., 2022). A possible explanation suggested to this disparity between models is their relative lysosome capacities (Carosi and Sargeant, 2023; Yuan et al., 2023). For instance, 5XFAD has less lysosomal degradation potential due to its mutant PSEN1 gene, which is known to interfere with the lysosomal biogenesis (Carosi and Sargeant, 2023; J. H; Lee et al., 2022). Faulty lysosomal degradation in 5XFAD and Tg2576 AD mouse models recently received support showing reduced lysosomal acidification, intra-lysosomal Aβ buildup, and cell death (J. H. Lee et al., 2022). Thus, it is reasonable that increasing autophagy flux via mTOR inhibitors may worsen Aβ response, although direct evidence has yet been demonstrated. The efficacy of mTOR inhibitor treatment in addressing Alzheimer’s disease (AD) pathology could also exhibit a significant temporal dependence, as illustrated by recent research findings. Specifically, Majumder and colleagues have shown that rapamycin effectively mitigates the accumulation of amyloid-beta (Aβ) plaques and tau tangles in the 3xTg-AD mouse model, but only if administered prior to the formation of these pathological features. In contrast, research by Lin et al. on the PDAPP(J20) mouse model suggests that inhibiting mTOR activity can arrest the progression of AD-like symptoms, diminish amyloid burden, and enhance cognitive function, even when the treatment is initiated post-disease onset (Lin et al., 2013). The observed discrepancies could be attributed to the genetic mutations present in various mouse models, presenting challenges in directly translating findings to human AD patients. Regrettably, this factor has not been adequately considered in the growing body of pre-clinical research that utilizes mTOR-associated inhibitors for Alzheimer’s disease treatment.

AD also entails the hyperphosphorylation and aggregation of tau, which is another significant hallmark of the disease. Under normal circumstances, tau is a soluble microtubule-associated protein that assists in microtubule formation and cytoskeleton stability. Tau hyperphosphorylation in AD forms insoluble tau structured termed paired helical filaments and NFTs. Diverse pathogenic effects are mediated by hyperphosphorylated tau, including inability for microtubule binding and microtubule network disassembly, aberrant signalling (Z. Cai et al., 2015), and protein aggregate accumulation (Ossenkoppele et al., 2022) to contribute to neuron death (Z. Cai et al., 2015).

AD patients that exhibit tauopathies often show mTOR activation. mTOR phosphorylation at Ser2448 and Ser2481 have been observed to colocalize with taupathy in the hippocampus and temporal cortex of AD patients (Griffin et al., 2005; J. H; Lee et al., 2022; X; Li et al., 2005). Moreover, mTOR effectors including GSK3β, S6K1, and 4E-BP1 also demonstrate increased phosphorylation at mTOR phosphorylation sites. Human genetic backgrounds that show activating mutations in the mTOR signalling pathway are also significantly more likely to develop tau hyperphosphorylation and AD (H. Y. Cai et al., 2023).

Similar to mTOR promoting Aβ oligomer formation, mTOR interacts with and integrates signalling pathways to propagate tau hyperphosphorylation and aberrant neurofilament formation (Extensively reviewed in (Mueed et al., 2018)). In brief, mTOR does not appears to directly phosphorylate tau but promote tauopathy via increasing tau-protein kinase activity and reducing its autophagic clearance. mTOR has been described to be upregulated by GSK3β, PI3K/Akt axis, insulin/IGF-1, and AMPK, often dysregulated in AD. This suppresses bulk autophagy clearance, and moreover, enhance Aβ-mediated tau hyperphosphorylation via tau kinases and tau CASP3 cleavage (X. Li et al., 2005). Corresponding to the feedback nature of mTOR and Aβ interactions, recent report show that intracellular tau accumulation also activates mTOR to further suppress autophagosome formation (M. Z. Li et al., 2022). Downstream targets of mTOR 4E-BP1 and S6K accelerate microtubule detachment of tau. S6K directly phosphorylates tau; 4E-BP1 release is suggested to mediate differential mRNA translation by eIF4E to impact tau (Ghosh et al., 2020). Indeed, mTOR upregulation in human SH-SY5Y significantly upregulated intracellular tau deposition in diverse cellular compartments (Tang et al., 2015). Interestingly, Tang et al. also showed that mTOR may also play a direct role in the subcellular localization of tau; mTOR mediated tau localization to exocytic vesicles. Lastly, interactions among mTOR, AMPK, and GSK3β also directly elevate tau phosphorylation at serine and threonine residues, hence promoting NFT formation. The roles of the mTOR kinase in Aβ oligomer formation and tauopathies are summarized in Figure 1.

Given these findings, the hypothesis that emerges is that inhibiting mTOR could slow down tau pathology progression in AD (Caccamo et al., 2010). Rapamycin, rapalogs and dual-mTORC1/2 blockers have all exhibited promising results in mitigating tau pathology across a variety of experimental models, as summarized in Table 2 (Caccamo et al., 2010; Frederick et al., 2015; T; Jiang et al., 2014a; Salama et al., 2018). For instance, the inclusion of 2.24 mg/kg rapamycin in the mice diet restored mTOR levels and alleviated tau pathology in 3xTg-AD mice (Caccamo et al., 2010). Rapamycin also reduced tau phosphorylation in rTg4510 mice when injected intraperitoneally (Morawe et al., 2022). It has been suggested that similar to pathological Aβ accumulation, tau pathology in AD is similarly temporally sensitive (Berron et al., 2021). Frederick et al. reported that both prophelactic and therapeutic temsirolimus treatment ameliorate taupathy in 3xTg-mice (Frederick et al., 2015). Rapamycin was administered at five or 10 months, with motor defect onset typically occurring at 8 months. Although total insoluble tau and phosphorylated tau concentrations were lowered in both treatments, NFT density in was only lowered upon prophylactic treatment at 5 months.

Studies on P301S tautransgenic mice largely support that mTOR inhibitors may attenuate tau hyperphosphorylation in AD when administered both therapeutically and prophylactically (T. Jiang et al., 2014b; Morawe et al., 2022; Ozcelik et al., 2013). Ozcelik et al., using the P301S model overexpressing the 0N4R tau isoform under the Thy-1.2 promoter element, observed a decrease in both total tau phosphorylation and tau tangles in the cortex upon rapamycin prior to and after tau hyperphosphorylation onset. Jiang et al., applying an unspecified P301S in the C57Bl/6J background, showed that therapeutic systemic temsirolimus treatment at 5 months reduces tau phosphorylation and improve spatial cognitive impairments. An upregulation of autophagy markers is a consistent theme, along with reduced 4E-BP1, S6K, and GSK3β phosphorylation. Cell line taupathy models substantiated these findings. SH-SY5Y neuroblastoma cells with established tau pathology reported a reduction in phosphorylated tau, autophagy induction, and GSK3β inactivation upon 100 nM temsirolimus treatment (T. Jiang et al., 2014a). In fact, mTOR inhibitors is reported to generally alleviate taupathies in cell models (T. Jiang et al., 2014a; Y; Liu et al., 2013; Silva et al., 2020). Recent evidence indicates that mTOR inhibition can limit tau phosphorylation associated with environmental causes of Alzheimer’s, including zinc (Lai et al., 2022) and aluminum toxicity (Y. Xu et al., 2022). Note that less commonly used mTOR inhibitors can also alleviate tau pathology. PP242, a dual mTORC1/C2 blocker, was reported to effectively reverse fenazaquin-established tauopathy (Salama et al., 2018).

However, a few studies registered inconsistent experimental outcomes. Using the 3xTg-AD mice, Majumder and colleagues only observed reductions in tau aggregation when rapamycin was given as a preventive measure, before the formation of Aβ plaques (Majumder et al., 2011). The effect of rapamycin on downstream proteins was markedly different when a different AD model was applied. Siman et al. established tauopathy model by injecting viral vectors for mutated human tau, containing four repeats and a P301L mutation, into CD1 background mice. The researchers observed that rapamycin suppressed various harmful effects of tau, including neuronal loss, synaptotoxicity, and gliosis, but the levels of human tau mRNA or protein were not altered (Siman et al., 2015). Recent investigation of therapeutic treatment of rapamycin in P301S mice, expressed under the Thy1-promoter, by Morawe et al. show that rapamycin administered therapeutically at 3.5 months does not reduce phospho-tau (T212/S214) levels or total insoluble tau levels (Morawe et al., 2022). It is conceivable that these effects may be due to varying pharmacological properties of mTOR inhibitors. Frederick et al. and Jiang et al. employ temsirolimus, which has greater solubility and delivery capacities pass the blood-brain barrier (BBB) than rapamycin, which Majumder et al. and Morawe et al. use. Morawe et al. show that PQR530, an PI3K/mTORC1 dual inhibitor with significantly greater BBB permeability than rapamycin, does alleviate tauopathy when administered at 3.5 months. Nonetheless, these disparities highlight the model-dependent nature mTOR inhibitors response. Furthermore, few studies evaluate the synergistic role of Aβ plaques in Alzheimer’s tauopathy, further urging clinical caution.

As in many neurodegenerative conditions, neuroinflammation represents a fundamental characteristic of Alzheimer’s disease. The activation of proinflammatory signalling cascades results in the release of immune mediators that profoundly affect neuronal functions, often culminating in cell death. Neuroinflammation has a dual impact on the brain. On one hand, it aids in clearing deposited Aβ, while on the other hand, it produces cytotoxic substances that exacerbate Aβ deposition and contribute to neurodegeneration. Numerous experiments have provided evidence of the high accumulation of inflammatory mediators around NFTs and amyloid plaques, highlighting the significant involvement of neuroinflammation in AD (Atri, 2019). The AD brain observed an increased level of microglia- and astrocytes-released IL-1, which facilitates the abnormal processing of APP and consequently the production of Aβ. Aβ, in turn, is considered to further boost the concentration of interleukins. Interleukins, including IL-4 and IL-10, affect Aβ clearance and inflammation. IL-4 decreases Aβ clearance, leading to the deposition of senile plaques, while IL-10 acts as an anti-inflammatory molecule, suppressing pro-inflammatory mediators. These findings highlight the intricate interplay between interleukins and AD pathology, where cytokine dysregulation contributes to neuroinflammation and neurotoxicity, ultimately demonstrating a clear correlation with AD development and progression.

The role of the mTOR in modulating immune responses is well-established, yet its specific influence on the development of AD through neuroinflammatory pathways is still a subject of ongoing research (Cheng, Wei, Qian, and Han, 2023). A compromised anti-inflammatory response can aggravate neuroinflammation, with several key signalling pathways—including NFκB, p38 MAPK, Akt/mTOR, nitric oxide, and COX—playing critical roles in driving immune reactions. Ko et al. found that the enhancement of autophagy via rapamycin treatment can inhibit the secretion of proinflammatory cytokines IL-1β and IL-18. This decrease in IL-1β correlates with a reduction in the activation of the p38 MAPK-NFκB pathway and a consequent downregulation of inflammatory agents like IL-6, IL-8, MCP-1, and IκBα in macrophages treated with rapamycin (Bagyinszky et al., 2017; Ko et al., 2017).

Moreover, the involvement of brain immune cells like astrocytes and microglia is significant, as their dysfunction is often linked to neurodegenerative conditions (Thakur et al., 2023). Microglia, as the resident macrophages and primary immune cells of the CNS, are essential in modulating inflammation and ensuring neuroprotection. Ye et al. found that lipopolysaccharide-induced inflammation reduces autophagic activity in N9 microglial cells, indicated by lower levels of the autophagy markers LC3-II and SQSTM1. Administration of the autophagy inducer, rapamycin, however, decreased the transcript level of pro-inflammatory cytokines, including TNF-α, IL-6, and iNOS (Ye et al., 2020). Similarly, astrocytes also facilitate normal brain function in a mTOR-dependent manner. Li et al. showed that by inhibiting mTOR signal pathway with rapamycin for 24 h, in astrocytes obtained from rat cerebral cortices, it suppressed oxygen-glucose deprivation/reoxygenation-induced astrocytic proliferation and migration. Concomitantly, following rapamycin administration, the expression of inflammatory mediators such as TNF-α and iNOS were reduced relative to control (C. Y. Li et al., 2015). This evidence collectively underscores the intricate interplay between mTOR signalling, glial cell regulation, and the neuroinflammatory landscape in AD.

In the context of Alzheimer’s models, mTOR inhibition has generally been associated with reduced brain inflammation and reduced glia activation. Figure 2 summarizes the known functions of the mTOR complexes in the brain. Oral administration of rapamycin to APOE4 transgenic mice starting at 1 month displayed reduced proinflammatory cyclophilin A (Cyp4) levels and cognitive function improvements (Lin et al., 2017). Inflammation induced by Aβ exposure may also be resolved by mTOR inhibition. Injection of Aβ25-35 into the hippocampus of C57BL/6 mice increased in vivo astrocyte proliferation, while simultaneous rapamycin and Aβ25-35 injection show significantly reduced proliferation (Y. C. Liu et al., 2017). At the same time, rapamycin administration reduced pro-inflammatory signalling markers, including NF-κBm, IL-1β, and TNF-α levels, of Aβ25-35-infused mice. (Y. C. Liu et al., 2017). These findings are consistent with other studies of mTOR inhibition in response to pro-inflammatory stimulants. Rapamycin reduces the pro-inflammatory environment associated with lipopolysaccaride (LPS) injections, physical damages to nervous tissue, and seizures through limiting various inflammatory cytokine release and astrocyte proliferation (Guo et al., 2017; C. Y; Li et al., 2015; F; Wang et al., 2023; Ye et al., 2020). We note that one study using intracerebroventricular (ICV) infusion demonstrated opposite results (X. Wang et al., 2016). Aβ1-42 injection followed by rapamycin showed an elevated inflammatory profile with increased IL-6 and TNFα levels, accompanied by increased pro-apoptotic caspase 3 activation and worsened spatial working memory ((X. Wang et al., 2016). ICV infusion is an uncommon method of Aβ1-42 and rapamycin delivery that bypasses the blood brain barrier, which may produce difference in effective dosage.

Whether a reduction in neuroinflammation following mTOR inhibition alleviates Alzheimer’s has been subject to mixed evidence. It has been found that reduced inflammation may reduce complement protein activation, free radical generation, cytokine release, and microglia and astrocyte proliferation and phagocytosis. Subsequently, reduced brain oxidative stress follows, which rescues amyloid plaque and tau tangle formation. Moreover, this reduces AD-associated nervous structure changes includes synapse engulfment and pyroptosis (Hong et al., 2016). This line of reasoning is supported by improved learning and memory capacities following anti-inflammatory treatments after LPS challenges or in response to other non-amyloidogenic neuroinflammatory conditions (Davoody et al., 2023).

Recent evidence show that rapamycin may suppress microglia activation and establishment of a microglia phagocytic profile. Shi et al. showed that activating mTOR kinases in the microglia of 5XFAD mice shows enhanced amyloid plaque clearance, improved cognitive functions, and reduced spine loss (Shi et al., 2022). Following rapamycin treatment, it reduced the number of mTOR-activated microglia and worsened amyloid pathology. mTOR activation increased phagocytosis receptor TREM2 expression and lysosomal biogenesis, which became suppressed upon mTOR inhibition (Shi et al., 2022). Reduced mTOR signalling is also associated with how amyloid-beta peptides can suppress microglia phagocytosis over time. Baik et al. show that chronic microglia exposure to Aβ reduces IL-1B and TNF-a expression and phagocytic activity in an AKT-mTOR-HIF1α signalling-dependent pathway (Baik et al., 2019). This desensitized phenotype has been reported to be analogous to the anti-inflammatory and non-Aβ phagocytic microglia profiles after mTOR inhibition, including the J20 model (Pomilio et al., 2020), 5xFAD isolated microglia (Grubman et al., 2021), and has been suggested in astrocytes. It is possible that the extent of Aβ-induced microglial exhaustion could account for the diverse response to rapamycin treatment in different neuroinflammation models.

The neurovascular system is a highly heterogeneous structure that involves diverse cerebral and microvasculature components (Muoio et al., 2014). This includes neurons, microglia, astrocytes, vascular endothelial cells, extracellular basement membrane, and a plethora of other cellular and acellular structures. They coordinate in response to external signals or local inducers to produce activities in a spatiotemporal specific manner, supporting both normal brain physiology and response to environmental stimuli. Examples include oxygen uptake, selective nutrient and waste exchange, neuronal signal transduction support, hormonal signal diffusion, etc.

A key function of the neurovascular system is responsible to maintain appropriate cerebral blood flow (CBF) quantities to support normal waste-clearance and oxygen nutrient distribution. The brain is well-known to consume 20% of total body energy supply, and requires an extensive oxygen delivery system for metabolism. This has been shown to be mediated by tight vascular cell dilation and constriction control via endothelial nitric oxide and neuronal nitric oxide production (Förstermann and Münzel, 2006). This dynamic signal integration allows for exact control of blood flow quantities in response to stimuli, through a process known as neurovascular coupling. This is responsible for rapid global or localised increases in nutrient uptake following elevated energetic demands from brain tissues. Disruptions in the process are commonly associated with disease (Förstermann and Münzel, 2006). Both reduced CBF (hypoperfusion) and excessive CBF (hyperperfusion) may cause acute ischemia, hypertension, stroke, and neurodegenerative diseases; neurovascular uncoupling is common to various forms of dementia.

Selectivity of nutrients and signals from the brain vasculature is maintained by the blood-brain barrier (BBB). The BBB is a highly complex structure which dynamically regulates molecule exchange between the cerebral interstitial fluid and blood (Reviewed in (Kadry et al., 2020)). Brain microvascular endothelial cells express transporters and junction proteins on the BBB for its activity and specificity. Moreover, recent evidence has demonstrated a key regulatory role by glial cells, notably microglia, astrocytes, and pericytes, in maintaining BBB integrity and function. In neurodegenerative diseases, physical damage to the BBB and protein expression alterations have been described to associated with reduced brain oxygen supply, reduced amyloid plaque clearance, neuroinflammation, neuron death, and gliosis.

Chronic cerebral hypoperfusion (CCH) has been noted to play an important role in the progression of AD through reducing extracellular Aβ plaque clearance, which may in turn further impair neurovascular activities, contributing to a self-sustaining cycle of disease progression. CCH is a key element of the vascular hypothesis, where cardiovascular diseases and amyloid plaque deposits are hypothesized to increase brain local inflammatory signals, elevate oxidative stress, compromise vascular functions, and induce neuronal cell death. Notably, CCH has been hypothesized as a primary mediator of Alzheimer’s progression that precedes tau hyperphosphorylation and amyloid plaque deposition (Park et al., 2019).

Current evidence largely supports mTOR inhibition to improve reduced CBF in AD. Interestingly, as opposed to worsened intracellular Aβ accumulation in late-stage lysosomal dysfunctions, mTOR inhibition has been consistently shown to rescue existing vascular defects and global Aβ levels. Lin et al. show that rapamycin administration at 7 months of age for 16 weeks to h (APP) mice reduced Anti-Ab antibody colocalization with brain vasculature, reduced microhemorrhages, restored CBF, and improved spatial memory (Lin et al., 2013). mTOR inhibition upregulates endothelial nitric oxide synthase (eNOS) to promote vasodilation and Aβ efflux from interstitial fluids (Lin et al., 2013). Notably, this indicates that late stage mTOR inhibition can improve AD pathologies via a non-autophagy mediated mechanism, in disagreement with lysosome-centric viewpoints as the acting mechanisms of mTOR inhibitors. Rapamycin treatment in the E4FAD Alzheimer’s mice model by Bell et al. replicates previous results, displaying increased CBF parameters, attenuated extracellular Aβ formation, and improved learning. The E4FAD model overexpress the human AD susceptibility allele APOE4, which may produce vascular damages years prior AD onset (Bell et al., 2012). Bell et al. administered rapamycin post-vascular defect onset, suggesting that mTOR inhibition by rapamycin may rescue both existing vascular defects and the following AB deposits. This may represent a body-wide improvement as opposed to only the cerebral vasculature. Rapamycin has also been recently shown to mitigate peripheral blood flow loss in h (APP) mice (Van Skike et al., 2023).

CBF restoration also leads to neurovascular uncoupling rescue. Neurovascular uncoupling occurs in Alzheimer’s patients where CBF does not increased in response to high neuronal activity. This cause oxygen and nutrient deprivation to directly cause neuron death and gliosis. Van Skike et al. fed rapamycin-supplemented chow to PDAPP(J20) mice at effectively 1.65 mg/kg weight, from 2 or 8 months to 10 or 12 months of age respectively (Van Skike et al., 2021). The control model displays significant loss in CBF increase upon whisker stimulation, indicating compromised neurovascular coupling. Both early and late rapamycin feeding, before and after the onset of neurovascular uncoupling, rescued this CBF reduction, reduced Aβ42 accumulation in the somatosensory cortex, and improved contextual memory following contextual fear conditioning (Van Skike et al., 2021). This temporally independent characteristic has been suggested to be due to the intact lysosomal system specific to J20 mice (Carosi and Sargeant, 2023). Distinct from their previous paper by Lin et al., the group demonstrates that this improvement is primarily mediated by mTOR elevation of neuronal NOS (nNOS) activity. When blocked with either the nNOS specific inhibitor L-NPA or general NOS inhibitor L-NAME, it erases the associated CBF improvements (Lin et al., 2013).

Recent findings by Jiang et al. show that rapamycin partially attenuates LRP-1 downregulation in human brain microvascular endothelial cell (HBMECs) models under high glucose conditions (G. Jiang et al., 2023). This reduces SREBP1 nuclear translocation and β-secretases expression, limiting Aβ synthesis. Rapamycin treatment in HBMECs and cerebrovascular endothelial cell raptor silenced mice both displayed increased Aβ efflux rates, and improved mice cognitive capabilities (G. Jiang et al., 2023). mTOR inhibition also reduces normal age-associated in microvasculature density declines and hypoperfusion, independent of Aβ (Van Skike et al., 2020). Of note, none of the studies described investigates the clearance of extracellular tau-tangles by mTOR inhibition’s effects on CBF. Emerging evidence from Hussong et al. show that taupathy blocks eNOS activation and induce vascular cell senescence (Hussong et al., 2023), which mTOR attenuation may effectively rescue given its overlapping role in both processes.

In BBB dysfunctions, current evidence largely points at mTOR inhibitors to restore brain BMEC protein expression changes associated with AD. In an h (APP)J20 mice model, rapamycin feeding at 5 months of age for 10 months reduced extravasated dextran levels in brain interstitial fluids and upregulated cell adhesion protein JAM-A (Van Skike et al., 2018). Rapamycin also downregulates matrix metalloproteinase-9 (MMP-9), an enzyme implicated in tight junction protein breakdown. The authors describe that mTOR inhibitors ameliorate BBB breakdown via restoring tight junction expression. However, the group did not find other BMEC tight junction proteins expression including claudin-5 or zonula occludens 1 (ZO-1) to change (Van Skike et al., 2018), suggestive of functional variations of tight junction proteins in AD pathology. mTOR inhibition also restores local vasculature homeostasis via BMEC efflux transporter upregulation. E4FAD mice following rapamycin treatment recovered p-glycoprotein levels to E3FAD control levels (Lin et al., 2020). P-glycoprotein is required for removal of toxic substances and xenobiotics, which become compromised in AD. We did not yet identify studies describing microglial inflammatory profiles in BBB maintenance and BMEC expression profiles under mTOR inhibition in AD, despite abundant associations of junction disassembly and global BBB integrity with microglia inflammatory states (You et al., 2016; Haruwaka et al., 2019; Huang et al., 2023), a process known to be controlled by mTOR.