mTOR in Alzheimer Disease and Its Earlier Stages: Links to Oxidative Damage in the Progression of this Dementing Disorder

Abstract

Alzheimer’s disease (AD) is the most prevalent form of dementia in the elderly population and has worldwide impact. The etiology of the disease is complex and results from the confluence of multiple mechanisms ultimately leading to neuronal loss and cognitive decline. Among risk factors, aging is the most relevant and accounts for several pathogenic events that contribute to disease-specific toxic mechanisms. Accumulating evidence linked the alterations of the mammalian target of rapamycin (mTOR), a serine/threonine protein kinase playing a key role in the regulation of protein synthesis and degradation, to age-dependent cognitive decline and pathogenesis of AD.

To date, growing studies demonstrated that aberrant mTOR signaling in the brain affects several pathways involved in energy metabolism, cell growth, mitochondrial function and proteostasis. Recent advances associated alterations of the mTOR pathway with the increased oxidative stress. Disruption of all these events strongly contribute to age-related cognitive decline including AD. The current review discusses the main regulatory roles of mTOR signaling network in the brain, focusing on its role in autophagy, oxidative stress and energy metabolism. Collectively, experimental data suggest that targeting mTOR in the CNS can be a valuable strategy to prevent/slow the progression of AD.

1. Introduction

Alzheimer disease (AD), the most wide-spread and prevalent form of dementia, is a neurodegenerative disorder affecting the elderly population. The number of people globally living with AD and other dementias has duplicated from 1990 up to date, mainly due to increased lifespan of the population. Neuropathologically, AD is characterized by the presence of two initial hallmarks: formation of senile plaques by extracellular deposition of amyloid β-peptide (Aβ) and neurofibrillary degeneration caused by aggregation of hyper-phosphorylated Tau protein. Further, progression is marked by selective neuronal death at selective vulnerable regions of the brain.

Amyloid β-peptide is a 39–43 amino acid peptide that undergoes misfolding and self-assembly in the form of oligomers, protofibrils, and fibrillar deposits in brains of AD patients and an earlier stage of this disorder, mild cognitive impairment (MCI). Aβ40 is the major Aβ species, whereas Aβ42, though being a minor species, readily aggregates into the above-listed aggregation states, and oligomeric Aβ42 is the more toxic of the two Aβ species. Both Aβ40 and Aβ42 are found in senile plaques [1]. Aβ is generated from the proteolytic cleavage of β-amyloid precursor protein (APP) through two divergent pathways [2]: the major (around 90% of APP) non amyloidogenic pathway normally involves the initial cleavage of APP by two membrane-bound proteases, α-secretase followed by γ-secretase; the minor - which is amyloidogenic - involves the action of β- and γ-secretases. Following its formation, Aβ can be degraded by proteolysis or cleared from the brain into the peripheral blood circulation.

The currently revised theory on Aβ-driven AD onset [3] has evolved from the consideration of amyloid fibrils as the predominant toxic form of Aβ towards smaller and soluble oligomers (AβO) or protofibrils [4]. This updated revision has gained increasing consensus based on the finding that the correlation between the numbers of senile plaques and the severity of dementia is poor, while correlation between small soluble AβO and degree of dementia is high.

The second protein associated with AD pathology is Tau, a structural protein that regulates the stability of tubulin assemblies that play a critical role in axonal transport [5]. Six tau isoforms are expressed in the adult human brain as a result of mRNA alternative splicing; 3R- and 4R-tau isoforms are located mainly in axons of adult neurons under normal physiological conditions. In AD brains, 3R and 4R tau is accumulated in a hyperphosphorylated state in pathological inclusions, referred to as neurofibrillary tangles (NFTs) [6]. Furthermore, Tau inclusions are thought to contribute to AD pathogenesis due to their occurrence in brain regions with altered function, and NFTs formation correlates with the duration and progression of the disease.

The clinical classification of AD is based on the severity of cognitive decline and the histopathological alterations [7]. Currently, five stages are described: i) preclinical: PCAD is characterized by by the absence of clinical signs and symptoms of AD (both typical or atypical phenotypes) and the presence of at least one biomarker of Alzheimer’s pathology [8] ii) amnestic mild cognitive impairment (MCI): characterized by lower levels of AD pathology, marked loss of memory, but normal activities of daily living [9]. The earliest pathological changes begin and are present in the entorhinal cortex and hippocampus iii) mild AD: cognitive symptoms are first manifested. The pathological alterations reach the cerebral cortex. Along with memory loss there is an inability to remember new information and disturbance in executive functioning. Subjects show personality changes, confusion and disorientation. iv) moderate AD: the symptom severity further increases. The pathological damage spreads to the areas responsible for language, reasoning and sensory processing (cerebral cortex). Behavioral problems, language disorder and impairment of visuo-spatial skills start to appear. v) severe AD: subjects completely lose their ability to execute daily activities. The pathological damage cover most of the cortical areas. Cognitive abilities significantly decrease, and systemic symptoms also appear (olfactory dysfunction, sleep disturbances, dystonia, and akathisia symptoms). AD pathology occurs approximately two decades before symptoms appear.

Other staging systems were developed for AD, including the Braak system [10], that is based on topographical staging of neurofibrillary tangles. This Braak classification categorizes AD progression in 6 stages and is an integral part of the National Institute on Aging and Reagan Institute neuropathological criteria for the diagnosis of AD [11].

In addition to the well-characterized plaque and tangle lesions, early deficits in synaptic function with later loss of neurons are considered to be crucial events in the onset and progression of AD. Both synapse and neuronal loss account for macroscopic cortical atrophy [12]. Increasing evidence shows that the loss of functional synapses represents an early defect in AD pathogenesis and that this synaptic loss correlates with memory impairment, even before evident neuronal loss [13]. Further, in AD NMDA and AMPA receptor-mediated excitotoxicity and defective compensatory GABAergic brain circuitry seem to occur concomitantly with synaptic dysfunction [14]. During early phases of neurodegeneration, AβO disrupt synaptic membranes resulting into a deficit in the long-term potentiation (LTP) accompanied by loss of memory and learning [15].

Synapse loss correlates with increased oxidative stress (OS) levels thus accounting for pathological changes observed during the clinically silent prodromal phases of AD [16]. Indeed, OS has been widely recognized as a prodromal factor associated with AD neurodegeneration [17]. Multiple factors mainly related to mitochondrial dysfunction and energy metabolism deficit contribute to elevate cellular OS. Brain structure, significant unsaturated lipid composition, high oxygen consumption, and rapid metabolic rate all contribute to brain susceptibility to toxic effects of OS. Moreover, the natural process of aging is associated with a physiological increase of OS over time [18, 19]. As an individual ages, ROS are integrated, to some extent, in the aging brain, disrupting redox-related communication and leading to cellular alterations, such as senescence and cell death, due to the inability to maintain redox homeostasis. This equilibrium in the cell is particularly important to keep the balanced microenvironment needed for multiple biological processes, from bioenergetics to vesicle transport and intracellular signaling. It is well-known that oxidative damage is greater in aged brains than it is in younger brains [20–23]. Notably, this “physiological accumulation” of oxidative damage with aging is exacerbated during neurodegeneration and is associated with loss of cognitive functions [24]. At the neuronal level, oxidative damage related to aging, strongly impairs the synaptic components involved in neuronal plasticity, cytoskeletal dynamics and cellular communication, among other alterations [25, 26].

In the context of AD neuropathology, several studies showed that OS can interfere with Aβ and Tau protein processing and functioning [27, 28]. Moreover, Aβ peptides themselves are involved in the production of ROS, which in turn causes mitochondrial dysfunction. The biomarkers of OS are well documented in the brain from AD patients. Among OS markers, enhanced levels of carbonylated proteins, mainly in parietal cortex and hippocampal region in brain, have been demonstrated by several authors [29–32]. Membrane proteins are more susceptible to undergo oxidation as compared to cytosolic proteins because radical reactions initiated at the membrane level produce highly reactive and neurotoxic lipid peroxidation products. Indeed, OS induces lipid bilayer modification, promotes lipid peroxidation thus increasing the levels 4-hydroxy-2-nonenal (4-HNE). In turn, protein modification can take place as a result of indirect oxidation by formation of protein adducts with 4-HNE. Increased levels of 4-HNE were reported in brain regions showing histopathological alterations of AD in this disorder and its earlier stage, MCI [31–36]. Oxidative modification of lipoic acid by 4-HNE was altered in AD hippocampus, with implications for pyruvate dehydrogenase and alpha-ketoglutarate dehydrogenase enzymatic activities [37]. In addition, the elevated protein carbonyl content in specific brain regions rich in senile plaques indicates a close association between oxidative damage and AD lesions [38]. Indeed, the brains of patients with MCI and AD also have increased oxidative alterations, such as protein nitration and nucleic acid modifications [39]. Fluorodeoxyglucose-positive emission tomography (FDG-PET) analysis revealed reduced brain glucose metabolism in MCI and AD, suggesting a role of mitochondria, the powerhouse of the cell [40].

In particular, redox proteomics studies from Butterfield’s laboratory allowed to identify the most susceptible proteins target of irreversible oxidative modifications, including carbonylation, nitration and HNE-adducts, reviewed in [41–43]. The group performed a comprehensive redox proteome analysis of different brain regions from late-stage AD (LAD), MCI subjects and also from subjects with EAD and PCAD [44, 45]. These results showed a map of specific oxidized proteins in the different stages of the disease. Collectively, these studies provided valuable insights into the molecular mechanisms involved both in the pathogenesis and progression of AD. The impairment of numerous cellular processes including energy production, cellular structure, signal transduction, synaptic function, mitochondrial function, cell cycle progression, and degradative systems are closely linked to oxidative damage to selected components of these pathways [46]. Considering that these cellular functions normally contribute to keep healthy neurons, loss of one or more of these functions could contribute to the pathology and clinical presentation of AD.

Further, the oxidation of nuclear and mitochondrial DNA has been reported in AD with amplified levels of oxidized bases like 8-hydroxyadenine, 8-oxo-2-dehydroguaninie and 5-hydroxyuracil inside frontal, parietal and temporal lobes [47]. Higher levels of 8-hydroxyguanine are detected inside hippocampus of patients with AD [48].

2. Protein oxidation, aggregation and degradation: a dangerous cocktail in the aging brain

Among toxic reactions initiated by ROS, oxidative modifications of proteins likely result in the loss of protein function, as discussed above. Further, protein carbonylation of hydrophobic amino acids buried in the structure of globular proteins lead to a significant dipole moment associated with the carbonyl moiety. This drives the carbonylated hydrophobic amino acids to the surface of the proteins, where the hydrophobic surfaces of different protein molecules aggregate in an attempt to exclude water, accounting for their aggregation [49, 50]. Overall, it is likely that free-radical mediated protein modification takes place in misfolding/aggregation process therefore accelerating dangerous protein deposition [51]. Proteins that are structured efficiently in the form of amyloid fibers or disorganized as amorphous aggregates and plaques are a common toxic aspect of these diseases. Though with diverse clinical manifestations Aβ, α-synuclein and Tau protein are key examples of chemical fingerprint underlying protein aggregation that has been demonstrated to strongly contribute to the pathogenesis and progression of neurodegenerative diseases [52].

However, in addition to protein misfolding and aggregation process, failure of protein quality control (PQC) mechanisms, such as chaperone or proteasome complex and autophagy, is a key pathological feature of age-associated neurodegenerative disorders [53]. When severely damaged, proteins cannot be refolded into their native shapes and the cell utilizes multiple security measure systems to maintain their functions. Chaperones will first act to rescue unfolded protein, and if unsuccessful, they can activate different cellular programs, including the unfolded protein response (UPR), heat shock response, ubiquitin-proteasome system (UPS), and endoplasmic reticulum-associated degradation to either solve the problem or eliminate the unfolded protein [54–56].

This latter drastic proteolytic removal of abnormal proteins is usually performed by the proteasome, a large macromolecular complex that degrades polyubiquitinylated, damaged proteins following removal of the poly(ubiquitin) chain, one residue at a time from the carboxy-terminal end employing the enzyme, ubiquitin carboxy-terminal hydrolase L-1 (UCHL-1). In AD, UCHL-1 is oxidatively dysfunctional [57, 58]. Alternatively, misfolded proteins with specific motif can be guided by HSP into the lysosome via translocation through the lysosomal-associated membrane transporter (LAMP2A), a process known as chaperone-mediated autophagy (CMA) [59, 60]. However, during aging, these systems may become less efficient and may be overwhelmed by the proteotoxic stress [61]. In particular, it is well-known that reduced proteasome activity may occur during aging and neurodegeneration, thus contributing to elevate the intracellular levels of oxidized/aggregated proteins [62, 63]. In AD, the UPS system is dysfunctional [64]. It is worth to be mentioned that at the same time oxidized/aggregated proteins may themselves inhibit the proteasome thus evading their removal [58, 65].

In parallel to proteasomal pathway, the lysosomal system is the most important “digestive mechanism” of the cell acting on different extracellular and intracellular substrates [60]. To do that, lysosomes, which are membrane-enclosed organelles, contain a variety of enzymes, including hydrolases, proteases, lipases, peptidases, glycosidases, nucleases, phosphatases, and sulfatases, able to digest all types of biological materials. Soluble, long-lived proteins, extracellular material, large protein aggregates, and even organelles can be targeted to the lysosome in a process called autophagy. The term autophagy derives from the Greek word “self-eating” and was coined in 1960s based on the observation of double-membrane vesicles entrapping organelles and proteins [66]. As a whole, autophagy is a self-degradative process that is crucial for balancing different sources of energy during starvation, both in development and in response to nutrient deprivation. Autophagy also plays a housekeeping role in removing misfolded/aggregated proteins, removing damaged organelles, such as mitochondria, endoplasmic reticulum and peroxisomes, as well as eliminating intracellular pathogens [67]. Thus, autophagy is generally considered a survival mechanism, though its aberrant regulation has been linked to non-apoptotic cell death.

The concerted function of UPS and autophagy may benefit cells by eliminating misfolded, dangerous proteins that accumulate and form intracellular aggregate structures in many major neurodegenerative diseases. Thus, understanding the molecular signaling signature that regulates both processes is essential to understand which event may contribute to disturbance of their correct equilibrium.

Among putative candidates, the mammalian target of rapamycin (mTOR) signaling pathway seems to be a key determinant in nutrient sensing and in regulating cell growth and autophagy [68]. The complexity of mTOR signaling, at the crossroad of multiple intracellular pathways, is discussed in the following sections, in particular focusing on the redox-mediated mechanisms in Alzheimer’s neurodegeneration.

3. Role of mTOR in physiology and in AD neuropathology

mTOR is a serine-threonine kinase that controls several important aspects of mammalian cell function. mTOR activity is modulated by various intra- and extra-cellular factors and its main function is to monitor whether cell resources and cell health are sufficient to respond to extracellular stimuli [69, 70]. Increasing evidence indicates that mTOR acts as a ‘master switch’ of cellular anabolic and catabolic processes, being involved in the regulation of a wide repertoire of cellular functions including cell growth and proliferation, cytoskeleton organization, and response to hypoxia. In the central nervous system (CNS), mTOR is significantly involved in synaptic plasticity, memory preservation, and neuronal recovery, while the dysregulation of mTOR is emerging as a common theme in a large number of human diseases, including cancer, metabolic syndromes and neurological disorders [69–71]. Recently, great attention has been placed on the role of mTOR in the development of AD since the alteration of mTOR activity was observed in AD brain and in AD mouse models thereof, supporting the notion that aberrant mTOR activity may be one of the leading events contributing to the onset and progression of AD hallmarks [72–77].

3.1. mTOR signaling

mTOR is a ubiquitously expressed protein complex, mainly localized in the cytoplasm whose catalytic activity is regulated by the phosphorylation at Thr-2446, Ser-2448 and Ser-2481 within the kinase catalytic (KIN) domains [69]. mTOR is known to be part of two different protein complexes: mTORC1 and mTORC2, which differ in some components, in upstream and downstream signaling, and in responsiveness to rapamycin treatment. Adjacent to the KIN domain is the FKBP12 rapamycin-binding domain, the site of inhibitory interaction between rapamycin and mTOR. Rapamycin disturbs mTOR–protein complex formation, thereby impairing mTOR activity. The mTORC1 core complex contains mTOR, regulatory associated protein of TOR (raptor), DEP-domain containing mTOR interacting protein (DEPTOR), proline–rich Akt substrate 40kDa (PRAS40), and mammalian lethal with sec-13 protein 8 (mLST8) [78]. In contrast, mTORC2 contains mTOR, mLST8, DEPTOR, rapamycin insensitive companion of mTOR (rictor), protein observed with rictor (Protor), and stress-activated protein kinase-interacting protein 1 (mSin1). The association of mTOR with either Raptor or Rictor determines the formation of mTORC1 or mTORC2. The heterodimer consisting of tuberous sclerosis 1 and 2 (TSC1 and TSC2) is a key upstream regulator of mTOR. TSC1/2 activation constitutively inhibit mTORC1/2 via Rheb (Ras-homolog expressed in brain) by stimulating the conversion of active Rheb-GTP to inactive Rheb-GDP. TSC1/2 activities rely on heterodimer formation and are controlled through by phosphorylation by a number of different kinases [70, 77, 79]. Among the upstream pathway that can activate mTORC1 by phosphorylation-dependent inhibition of TSC1/TSC2 complex is worth to mention PI3K/Akt and Ras/Erk, which are activated by growth factors like insulin or Akt which also inhibit PRAS40 by phosphorylation [70, 80–83]. A primary inhibitory-feedback loop exists among these pathways whereby sustained activation of mTOR/p70S6K suppresses AKT activity.

The above-described mechanism acts through mTORC1-mediated phosphorylation of IRS-1 serine, to induce IRS-1 inactivation and degradation, thus eliminating the coupling of PI3-K/Akt to insulin and IGF-1 receptors. This process results in and is a major cause of insulin resistance [84–87]. mTORC2 phosphorylates Akt to activate the pathway, but also promotes Akt degradation. Similarly, JNK, TNFα and other inflammatory molecules antagonize and turn off PI3-K/Akt axis, deregulating IRS-1 activity. In addition, Parkin7 (DJ-1), under acute oxidative stress, can activate the ERK pathway, which increase mTORC1 activity by directly inhibiting TSC1/2, or by indirectly interacting with p70S6K [77].

mTORC1 is involved in the response to cellular energy demand through AMPK. During low cellular energy, AMPK inhibits mTORC1 through the direct phosphorylation of raptor, which disrupts the mTORC1 complex, and through the phosphorylation of TSC1/2 [81, 88]. Moreover, increasing cellular amino acid or glucose concentration activates mTORC1 through the Rag family of GTPases. Growth factors can activate mTORC2 via the PI3K signaling pathway, increasing the association between mTORC2 and ribosomes. mTORC2 is involved in cellular metabolism and cell shape interacting with actin through PKC-α and Rho GTPase, and in cellular proliferation regulating AKT. mTORC2 is also modulated by increased glucose though rictor acetylation [69, 70, 89].

p70S6K and 4EBP, which regulate protein translation, and Ulk1 and Atg13, which suppress autophagy, are the best-characterized substrates for mTORC1. During availability of high cellular energy resources, mTORC1 promotes ribosome biogenesis and mRNA translation through phosphorylation and activation of p70S6K and 4EBPs. mTORC1- activated p70S6K phosphorylates substrates to promote translation and elongation (i.e., 40S ribosomal subunit proteins, eiF4B and eEF2K), while mTORC1-phosphorylated 4EBPs promote cap-dependent translation through their participation in the eIF4F complex. mTORC1 also regulates lipid biosynthesis through activation of SREBP1, a major lipogenic transcription factor that controls genes involved in fatty acid and cholesterol synthesis [90–92]. In physiological conditions mTORC1 directly interacts with the Ulk1 complex (Ulk1-Atg13-FIP200-Atg101), thus controlling autophagy at the stage of phagophore formation. Autophagy allows the clearance of obsolete proteins, misfolded or overexpressed protein aggregates, and whole organelles [69]. The mTORC1 complex phosphorylates and inhibits Ulk1 and its interacting partner Atg13 as well as AMBRA1, the key link between Ulk1 and Beclin-1 complexes. Under starvation conditions or rapamycin treatment, mTORC1-mediated phosphorylation of Atg13 and Ulk1 is inhibited, leading to dephosphorylation-dependent activation of Ulk1 and Ulk1-mediated phosphorylation of Atg13, FIP200, and Ulk1 itself that triggers autophagy initiation [82, 93, 94]. Furthermore, nutrient deprivation or mTORC1pharmacological inhibition promote the induction of the energy-sensing kinase AMPK, which interact and activate Ulk1, but also directly inhibit mTOR through phosphorylation of TSC2 [95]. Recent studies demonstrated that mTORC1 regulates lysosomes through the transcription factor EB (TFEB), which controls many genes whose associated proteins play key roles in lysosomal function. mTORC1 inhibition favor the nuclear accumulation of TFEB and thus its activity, promoting autophagy induction when nutrient levels are low [96].

3.2. mTOR in brain development and surveillance

Although the precise mechanisms are still not fully understood, regulated and coordinated activities of mTORC1 and of mTORC2 are necessary for the development of CNS [69]. During embryonic development, mTOR signaling functions as a powerful neuronal survival and division signal in response to growth factors and guidance cues. In brain maturation, the mTOR pathway promotes extension of neurites (dendrites and axons) and, through the induction of protein and lipid synthesis, increases cellular mass with expansion of plasma membrane, thus favoring axon guidance [97, 98]. mTORC2, by affecting actin dynamics, may be a putative facilitator of growth cone motility, including neurite path finding and elongation. The same mTOR-mediated regulatory mechanisms that control division/migration programs during brain development evolve in adult individuals to control synaptic plasticity, neuronal polarity, neurotransmission, metabolic control, proteostasis and stress responses. Therefore, in post-mitotic neurons the mTOR pathway has significant functional impact on synaptic processes underlying memory and learning. mTOR is crucial in almost all kinds of neuronal plasticity (e.g., LTP, LTD) and different learning and memory processes [70, 77, 99]. mTOR activity also contributes to nervous system regeneration. The activity of mTOR is essential in terms of normal cognition, whereas hyperactivity of mTOR can be harmful to the brain functions [79]. Furthermore, mTOR is also important for feeding and body weight control, circadian rhythm, and nociception [70].

In aged brain mTOR/autophagy axis plays a dual role in the cellular response to stress and in the regulation of the proteostasis network [90–92]. Increasing evidence suggests that reduced autophagy could lead to accumulation of toxic aggregates and oxidized proteins [56, 100–102]. In addition, the age-dependent reduction in autophagy is responsible of the build-up of severely damaged mitochondria, thus exacerbating oxidative stress and tissue damage with age. In agreement with its role in promoting cell survival and increasing life span during physiological conditions, a deregulated mTOR/autophagy axis under stress conditions (e.g., oxidative stress) may result in cell death [103–105].

3.3. mTOR in the development of AD

In the last two decades mTOR signaling has been extensively analyzed in AD brains and in brains from AD mouse models, demonstrating its aberrant up-regulation during neurodegeneration [74, 106–108]. Data from AD brains indicate that mTOR phosphorylation (both Ser-2448 and at Ser-2481) and the regulation of its downstream targets, p70S6K and eIF4E, are increased in the hippocampus and in other brain areas [109–113]. In addition, mTOR hyperactivity correlated with Braak stages and/or cognitive severity of AD patients. Increased mTOR activity was also observed in the brain from MCI individuals but not in preclinical AD (PCAD) subjects [74]. Further, the dysregulation of the mTOR signaling was observed in peripheral lymphocytes of patients with AD, which correlated with the progression of the disease [114].

Consistent with mTOR data, several studies demonstrated the impairment of the PI3K/Akt axis in AD brain as well as ERK1/2 signaling. The sustained activation of neuronal PI3K/Akt/mTOR pathway in AD and MCI brain and in AD models was associated with insulin receptor substrate 1 (IRS1) inhibition, disabling normal activation of PI3K/Akt by insulin and inducing insulin resistance [74, 87, 103, 115–117]. The hyperactivation of the PI3K/Akt/mTOR axis is also associated with reduced autophagy in AD brain leading to the accumulation of protein aggregates. Several markers of autophagy initiation and execution or autophagosome formation (e.g., LC3 II/I) and of autophagic flux are reduced in AD brain and in AD early stages [92, 97, 113].

Consonant with the above discussion, pharmacological reduction of mTOR hyperactivity in the brains of AD and of AD-like mouse models was shown to restore autophagy [118–122]. The reduction of autophagy in AD also was shown to be influenced by a number of different events, such as increased OS, but not directly related to mTOR hyperactivation [60, 73, 102, 123]. The deterioration of neuronal autophagy, observed in AD, most likely allows increased Aβ production and aggregation, and contributes to the malfunction of Aβ elimination from the brain. Indeed, the autophagy-lysosome pathway is an important regulator of APP processing and clearence and also one of the major cellular pathways for the removal of Aβ aggregates [56, 91, 92, 103, 107, 123, 124]. In addition, the analysis of the brain from AD subjects and mouse models of the disease indicated a strong link between increased mTOR signalling and tau neuropathology comprising the formation of tau oligomers and insoluble aggregates [74, 75, 112, 125]. In a small number of studies concerning the analysis of mTOR signaling in AD animal models contrasting results were shown. Indeed, the analysis of APP/PS1 double transgenic mice and of Tg2576 mice demonstrated the inhibition of mTOR signaling, thus hypothesizing that mTOR alterations results might depend on the strain of the models analyzed and/or on the age of the mice [126, 127].

4. The crosstalk between mTOR and mitochondria in AD

Mitochondria play an essential role in energy generation, cell signaling, differentiation, death, and senescence in eukaryotic cells. In particular, the role of mitochondria in cellular senescence and neurodegeneration is widely associated with their significant function as generators of ROS- mediated random molecular damage. Indeed, ROS have been shown to induce genomic damage [128], accelerate telomere shortening [129], impair actin dynamics and synaptic plasticity mechanisms [130, 131], and to act as drivers of signa ling networks important for the maintenance of the senescent phenotype [132]. Accumulation of ROS such as hydrogen peroxide (H₂O₂) is an oxidative stress response, which induces various defense mechanisms or programmed cell death [103, 125, 133, 134]. As one of the major types of programmed cell death, autophagy has been observed in response to several anticancer drugs and demonstrated to be responsible for cell death. However, the exact mechanism by which ROS regulates autophagy is still poorly understood. The role of mTOR appears crucial in regulating mitochondrial functions and ROS production. Baker et al. [135] proposed that mitochondria and their capacity to generate free radicals near mitochondrial DNA (mtDNA) are one of the most studied possible mechanisms of aging. In addition to that, Correira-Melo et al. proposed a mechanism by which the DNA damage response, a major driver of senescence, interacts with mitochondria to develop a senescent phenotype. In particular, Akt and mTORC1 phosphorylation cascade would play a key role as an effector of the DNA damage response. Indeed, mTOR activity was decreased in senescent mitochondria- depleted cells [136].

4.1. mTOR, mitochondria and oxidative stress

Many investigations have consistently shown that all the interventions that increase longevity in mammals, i.e., dietary, protein or methionine restriction are associated with reduced mTOR activation, decreased mitochondrial ROS production, and decreased oxidative-derived damage possibly contributing to longevity extension and reduced neurodegeneration [135]. Moreover, as reported by our group both in AD and DS brain the accumulation of protein oxidative damage results from the increased free radical production, mainly related to metabolic alterations, mitochondrial degeneration, and amyloid-β oligomers, and aberrant activity of protein degradative systems, including mTOR-dependent autophagy [72, 74, 78, 80, 84, 137]. In this context, the early aberrant hyper-phosphorylation of mTOR coupled with the reduction of autophagosome formation was found to be associated with increased levels of 3-NT and 4-HNE Ts65Dn mice (a well-known model to study DS), suggesting the potential involvement of altered autophagy in the buildup of protein oxidative damage [138, 139]. Data obtained in vitro strengthened the protective role of autophagy in reducing protein oxidation [138].

In addition, concurrent phosphorylation of AMPK and mTOR was observed in AD brains with high colocalization with hyperphosphorylated Tau [140]. Mitochondrial antioxidant enzymes e.g., SOD2, p1, and p4 were substantially decreased in p-AMPK, p-mTOR, and p-tau positive cells along with higher levels of DNA and protein oxidation [140]. Similarly, a novel molecular mechanism favoring the disruption of the AMPK/mTOR axis and the impairment of autophagy along with the accumulation of oxidatively-damaged proteins and lipids within the brain was recently reported [81]. Together, these lines of evidence suggest that AMPK and mTOR metabolic axis has a crucial role in AD brains and that their dysfunction leads to increased oxidative stress levels in AD. As originally proposed by our group [32, 74] and expanded by Kim et al. [141], Aβ peptides may be crucial determinants linking ROS, mitochondria, and mTOR in the AD brain. These authors showed that Aβ treatment leads to increased ROS generation and mitochondrial fission, accompanied by dysfunction of mitochondria such as loss of membrane potential and ATP production in neuronal cells [141]. These events were coupled with the sustained Akt activation, that induced not only the fragmentation of mitochondria but also the activation of mTOR, eventually suppressing autophagy. Indeed, inhibition of autophagic clearance of Aβ led to increased ROS levels and aggravating mitochondrial defects, which were blocked by rapamycin (noted above to be a mTOR inhibitor). Hence, sustained phosphorylation of Akt by Aβ directly inhibits autophagy through the mTOR pathway and these changes elicit abundant mitochondrial fragmentation resulting in ROS-mediated neuronal apoptosis [141].

Among the strategies to clarify the role of mTOR hyperactivation in mediating mitochondrial defects and increased oxidative damage leading to brain dysfunctions and AD, rapamycin (an mTOR inhibitor) administration appears to be of great interest. While the number of studies addressing the beneficial effects of rapamycin in neurodegenerative disorders is consistently increasing in the last years, only a limited number of papers, including those from our group, addressed a potential role for rapamycin in ameliorating mitochondrial defects in AD and are discussed in the current review. Rapamycin administration promoted autophagy and markedly reversed Aβ1–42-induced impaired redox homeostasis by decreasing the levels of prooxidants-ROS generation, intracellular Ca²⁺ flux, and lipoperoxides, and increasing the levels of antioxidants, i.e., SOD, catalase, and GSH in adult rats [82]. Autophagy activation following rapamycin treatment also provided significant neuroprotection against Aβ1–42-induced synaptic dysfunction by increasing the expression of synapsin-I, synaptophysin, and PSD95; and neurotransmission dysfunction by increasing the levels of CHRM2, DAD2 receptor, NMDA receptor, and AMPA receptor; and ultimately improved cognitive ability in rats [82]. Moreover, rapamycin abolished the age-associated increase of mitochondrial ROS production and percent free radical leak at complex I, accumulation of mtDNA fragments inside nuclear DNA, mitochondrial protein lipoxidation, and lipofuscin accumulation in mice. Rapamycin also decreased RAPTOR and increased PGC1-α and ATG13 proteins. These findings support the possibility that a decrease in the rate of free radical generation at mitochondria and increased autophagy coordinately contribute to the increase in longevity and to reduced neurodegenerative processes induced by rapamycin in aged CB6F1 mice [94]. Our group also reported that intranasal administration of rapamycin to target the brain ameliorate Alzheimer-like cognitive decline and reduce the oxidative damage in the brain of a mouse model of DS, by reducing mTOR hyperactivation [118, 137].

An interesting observation comes from the work performed by Till et al. [142], highlighting a role for GABA in mediating mTOR activation, autophagy impairment, and mitochondrial defects in the brain. Indeed, autophagy operates specifically to degrade particular proteins or organelles, such as peroxisomes (pexophagy) [142], mitochondria (mitophagy) [143], or ribosomes (ribophagy) [144]. Much of the same core machinery used for general autophagy also overlaps in the selective autophagy pathways. In this picture, GABA was proposed as a regulator of mitophagy and pexophagy via mTOR activation. When GABA levels are increased, cells are unable to specifically degrade mitochondria and peroxisomes via these selective autophagy pathways during starvation conditions, leading to oxidative stress due to the accumulation of these organelles. Indeed, GABA administration leads to an increase of ROS, which can be prevented by co-treatment with glutathione, but even more by rapamycin, thus suggesting that GABA increases cellular oxidative stress, probably due to the presence of longer- lived or damaged peroxisomes and mitochondria because impaired autophagy [145].

Also, lysosomes are essential for autophagy, and autophagic clearance of dysfunctional mitochondria represents an important element of mitochondrial quality control. Indeed, alterations in lysosomal function can affect upstream autophagic functions. Loss of ATP13A2, which encodes a lysosomal P-type ATPase, was shown to affect mitochondrial function through mTOR [146]. Knockdown of ATP13A2 led to an increase in mitochondrial mass in primary mouse cortical neurons and in SH-SY5Y cells forced into mitochondrial dependence. ATP13A2-deficient cells exhibited increased oxygen consumption without a significant change in steady-state levels of ATP. Mitochondria in knockdown cells exhibited increased fragmentation and increased production of ROS. ATP13A2 knockdown cells exhibited decreased autophagic flux, associated with increased levels of phospho-mTOR, and resistance to autophagy induction by rapamycin. The effects of reduced ATP13A2 levels on oxygen consumption, mitochondrial mass and ROS production are mimicked by inhibiting autophagy induction. Hence, hyper-active mTOR-dependent decreased autophagy associated with ATP13A2 deficiency impairs mitochondrial quality control, resulting in increased ROS production [146]. These data, therefore, suggest that lysosomal impairment affects mitochondrial homeostasis further supporting the evidence for mitochondrial dysregulation in AD neuropathology [103]. In vitro, rapamycin inhibited S6 phosphorylation events, mitochondrial ROS production in hippocampal neuronal cell line HT22, thus suggesting a role for mTOR in regulating mitochondrial bioenergetics in the brain [83].

4.2. mTOR, mitochondria and energy metabolism

Human aging is characterized by a gradual reduction in the ability to coordinate cellular energy expenditure and storage (crucial to maintain energy homeostasis), and by a gradual decrease in the ability to mount a successful stress response [147, 148]. These physiological changes are typically associated with changes in body composition (i.e., increase in fat mass and the decline in fat-free mass), and with a chronic state of oxidative stress with important consequences on health status [149–151]. In particular, mitochondria have a crucial role in the supply of energy to the brain. Mitochondrial alterations other than an increased ROS production can lead to detrimental consequences to the function of brain cells and for that reason are thought to have a pivotal role in the pathogenesis of several neurologic disorders. Within the brain, mitochondrial functions deputed to energy production are needed to regulate synaptic transmission, neurotransmitter recycling, dendritic and axonal transport, ion channels, and ion pump activity, which are processes with high energetic requirements [152]. In this sense, mitochondrial bioenergetic failure has been pinpointed as a mechanistic event underlying brain malfunction, senescence and/or neurodegeneration [103, 153].

Redox proteomics techniques in AD brain allowed the identification of specific mitochondrial proteins, including some in the mitochondrial oxidative phosphorylation machinery, that are targets of oxidative modifications during neurodegenerative processes and thereby become dysfunctional thus impairing mitochondrial bioenergetics [134]. ATP synthase alpha, manganese superoxide dismutase (MnSOD), malate dehydrogenase and VDAC were identified modified by both protein carbonyls, 3-NT and HNE-bound, in the early stages and in the late stage of AD, in MCI and EAD [134, 154]. Indeed, Eno1 and ATPase activities are reduced in AD brain [155]. ATP synthase, the last complex of the electron transport chain (aka complex V), is an essential enzyme in the inner mitochondrial membrane and plays a key role in energy metabolism. Its oxidative modifications modify its conformation leading to the inactivation of the complex. The defective function of ATP synthase activity secondary to oxidative modification significantly contributes to lower ATP levels, possibly resulting in electron leakage and increased ROS production [45]. Dysfunction of single complexes of the respiratory system is frequently accompanied by deleterious side effects like loss of mitochondrial membrane potential (MMP) and consequently decreased ATP levels, but also the production of ROS [156]. Dysfunction of single enzyme complexes, ROS production, mitochondrial permeability transition pore (mPTP) opening, elevated apoptosis, in addition to structural alterations, and a diminished mitochondrial content are believed to critically contribute to the onset and progression of neurodegenerative pathology in AD [157–160].

Consonant with the above, deficiency in several mitochondrial key enzymes is well documented in AD. These include enzymes involved in the TCA cycle, such as aconitase, ketoglutarate dehydrogenase complex and pyruvate dehydrogenase complex as well as those involved in the electron transport chain of oxidative phosphorylation such as cytochrome oxidase [41]. Targeting brain metabolism and mitochondrial function are relevant to hypometabolism and impaired mitochondrial bioenergetics that are among the earliest pathogenic events in AD pathology [134]. While the involvement of mitochondrial failure in AD has been extensively reviewed, also by our group [161], in the current review we focus on the involvement of mTOR in regulating mitochondrial bioenergetics in AD. In particular, in light of the fact (as described in the section above) that pharmacological agents that inhibit mTOR can protect neurons against dysfunction and degeneration in animal models of acute brain injury and neurodegenerative disorders. Therefore, a better understanding of such adaptive responses of neurons to bioenergetic challenges may lead to the development of novel approaches for promoting optimal brain function and for preventing and treating neurodegenerative disorders.

Among the processes identified to impair mitochondrial bioenergetics favoring increased oxidative stress during aging and AD the dysregulation of uncoupling proteins (UCPs) emerge as a pivotal event. In humans, UCPs are a group of five mitochondrial inner membrane transporters with variable tissue expression, which seem to function as regulators of energy homeostasis and antioxidants [162, 163]. In particular, these proteins uncouple respiration from ATP production, allowing stored energy to be released as heat. Data from experimental models have previously suggested that UCPs may play an important role in the rate of aging and consequent lifespan. Mitochondrial uncoupling imposes an energetic stress on cells by causing a proton leak across the inner membrane which reduces the membrane potential and uncouples substrate oxidation from ADP phosphorylation, thereby increasing energy expenditure [162, 163]. Mitochondrial uncoupling is a physiologically regulated process that plays important roles in the adaptive responses of organisms to changing environmental conditions. Mitochondrial uncoupling proteins regulate multiple physiological processes including thermogenesis, mitochondrial redox balance and free radical production, cellular calcium homeostasis, and autophagy/mitophagy. Thus, by promoting fatty acid oxidation, and by reducing ATP and ROS production, the induction of mitochondrial uncoupling through UCPs may also be a critical pathway in the modulation of the rate of aging and lifespan [164–166].

A potential role for mTOR in regulating UCPs was proposed by Mattson and co-workers [167]. These authors showed that mild mitochondrial uncoupling enhances cellular energy expenditure in mitochondria and can be induced with 2,4-dinitrophenol (DNP), a proton ionophore previously used for weight loss. DNP treatment reduces mitochondrial membrane potential, increases intracellular Ca²⁺ levels and reduces oxidative stress in cerebral cortical neurons. Gene expression profiling of the cerebral cortex of DNP-treated mice revealed reprogramming of signaling cascades that included the suppression of mTOR and insulin-PI3K-MAPK pathways, and up-regulation of tuberous sclerosis complex 2, a negative regulator of mTOR [88]. Genes encoding proteins involved in autophagy processes were up-regulated in response to DNP. CREB signaling, Arc and brain-derived neurotrophic factor, which play important roles in synaptic plasticity and adaptive cellular stress responses, were up-regulated in response to DNP, and DNP-treated mice exhibited improved performance in a test of learning and memory [167]. Hence, these findings suggest that mild mitochondrial uncoupling triggers an integrated signaling response in brain cells characterized by reprogramming of mTOR and insulin signaling, and up-regulation of pathways involved in adaptive stress responses, molecular waste disposal, and synaptic plasticity.

Physiological bioenergetic challenges such as exercise and fasting also can enhance neuroplasticity and protect neurons against injury and neurodegeneration. Conversely, in AD brains the UCPs gene expression decreased significantly relative to control brains, and the mean levels tended to be lower in AD brains with higher grades of neurodegeneration. Considering that mTOR hyperactivation has been reported in AD and its earlier stage, MCI [74], we speculate about a possible involvement for mTOR in repressing UCPs expression. Overall, the failure to maintain normal levels or to increase the expression of UCPs may potentiate oxidative stress leading to progressive mitochondrial DNA damage and energy depletion [168]. As recently reported by Montesanto et al. [169], rs9472817-C/G, an intronic variant of neuronal mitochondrial uncoupling protein-4 (UCP4/SLC25A27) gene affects the risk of late-onset Alzheimer’s disease (LOAD), and that the variant’s effect is strongly dependent on APOE-ε4 status [169].

Another pathway including mTOR as a key player and involved in the regulation of mitochondrial bioenergetics that was found to be impaired in AD is the insulin signaling cascade [85]. Insulin signaling activation triggers multiple effects, including synthesis of proteins involved in neuronal glucose metabolism, anti-apoptotic mechanisms, and antioxidant defense [85, 170]. It is worthy of mention that insulin (a) prevents the disruption of mitochondrial membrane potential and cytochrome c release, thus promoting neuronal survival; and (b) enhances mitochondrial membrane potential, ATP levels, nicotinamide adenine dinucleotide phosphate (NADPH) redox state, and hexokinase activity, thus favoring mitochondrial bioenergetics [85, 170]. In light of these roles, the development of brain insulin resistance was proposed among the mechanisms responsible for increased oxidative damage in the brain [43, 139, 171] as well as for the development of AD [85, 172]. Brain insulin resistance mediates synaptotoxic effects in several ways leading to (1) synapses loss, impaired autophagy and increased neuronal apoptosis [85, 115, 173]; (2) increased production and secretion of beta-amyloid peptides (Aβ) [85, 174]; as well as (3) an increased Tau phosphorylation [85, 125, 175]. These observations collected in vitro and in AD mouse models have been also strengthened by clinical studies reporting that the failure in brain energy metabolism responsible for cognitive decline during aging or AD could be driven by the development of brain insulin resistance particularly at early stages [85, 176]. On the other hand, strategies to overcome brain insulin resistance, such as intranasal insulin administration, were shown to ameliorate memory and cognitive functions [170]. Similar results were obtained with rapamycin [118, 137] treatment, or by genetically reducing mTOR hyperactivation [86, 177], suggesting a role for mTOR in the metabolic effects mediated by insulin in the brain. Taken together, these findings support a crucial role for mTOR in mediating brain insulin effects on several processes that are essential for healthy aging.

Aberrant mTOR phosphorylation was proposed to be a key driver for IRS1 inhibition and brain insulin resistance [85, 103, 178, 179]. Indeed, increased mTOR activation and IRS1 inhibition were reported in both MCI and AD brain [74] as well as in mouse models of AD [103, 171, 174, 180]. Among the pathways responsible for mTOR hyperactivation in AD, our group identified the impairment of biliverdin reductase-A as a crucial event responsible for the disruption of the AMPK/mTOR axis resulting in mTOR hyper-activation in the brain [81, 84, 171]. Within the insulin signaling pathway, mTOR hyper-activation would be responsible for mitochondrial damage and reduced energy production. As reported by Swerdlow and co-workers [159], hybrid cells, modeled by transferring mitochondria from MCI, AD and control subject platelets to mtDNA-depleted SH-SY5Y cells, were characterized by altered mTOR along with an altered mitochondrial dynamic and bioenergetics, as well as, reduced glycolytic flux [181]. Moreover, Norambuena et al. [182] showed that Aβ oligomers (AβOs) activate mTORC1 at the plasma membrane in neurons, which leads to cell cycle re- entry, a frequent prelude to the death of neurons [87]. In doing so, AβOs-mediated mTOR hyperactivation also interferes and suppresses the nutrient- induced mitochondrial activity, thus generating a sort of prolonged starvation phenotype in AD neurons. As noted above, AβOs in the membrane bilayer led to lipid peroxidation with production of neurotoxic HNE [183]. Therefore, AβOs initiate two parallel, interconnected pathways that together lead to neuronal dysfunction in AD [182]. Rather, this effect of AβOs reportedly can be blocked by insulin and nutrients, like amino acids [87].

Finally, a significant decrease in the protein levels of LC3-II and a significant increase in protein levels of mTOR phosphorylated at serine residue 2448 was observed in GK rats (a model for type 2 diabetes mellitus) suggesting suppression of autophagy in the diabetic brain cortex. No significant alterations were observed in the parameters related to mitochondrial biogenesis. Altogether, these results demonstrate that during the early stages of T2D, brain mitochondrial function is maintained in part due to a delicate balance between mitochondrial fusion-fission and biogenesis and autophagy [184]. Moreover, these events are precedent to an age-related impairment of the respiratory chain and uncoupling of oxidative phosphorylation documented in brain mitochondria isolated from 12- to 24-month GK rats [185].

4.3. mTOR, mitochondria and lifespan

Aging is the major risk factor for developing AD and is the most complex phenomenon in nature, where an inevitable time-dependent, progressive functional decline of diverse physiological functions finally results in death [103]. Aberrant mitochondrial function is a major characteristic of aging cells and is thought to contribute to neurodegeneration observed in AD [186]. Remarkably, genetic models of mitochondrial dysfunction (such as mutation of mitochondrial DNA polymerase in mice) correlate with reduced life span [187]. Surprisingly, many interventions that extend life span, e.g., caloric restriction and rapamycin, by modulating mTOR activity lead to reduced energy intake and decreased mitochondrial functions [135]. These findings revealed complex and antagonistic functions of mitochondria in aging and have been at least partially reconciled by biphasic modeling of mitochondrial dysregulation [135]. This model proposes that alterations of mitochondrial functions are not linear with aging of the organism, but rather that they increase and peak in middle age, followed by a decline in older age [135]. Mitochondrial dysfunction in aging cells is characterized by several factors, including elevated ROS production and mitochondrial DNA mutations; decreased electron transport function, membrane potential, and ATP production; altered mitochondrial dynamics; or dysregulated mitophagy [186, 188]. Functionally, altered mitochondrial activity participates in inducing cellular senescence [189], chronic inflammation, and a decline in stem cell activity associated with aging [187]. Nevertheless, while excessive ROS production is detrimental for neurons, low ROS levels could promote the activation of AMPK through S-glutathionylation of reactive cysteines located at the α- and β-subunits of AMPK [102]. This modification would further lead to phosphorylation and activation of the tuberous sclerosis complex1 (TSC1)/TSC2 and consequent inhibition mTOR activity [88].

This mechanism seems to be at the base of the interventions that slow aging and prevent chronic disease, including caloric restriction (CR). CR impacts on the major nutrient-sensing pathways that are mTOR, AMPK, sirtuins and insulin/IGF-1, which sense the availability of macronutrients (glucose, amino acids and lipids) or the energy status of the cell (AMP/ATP or NAD⁺/NADH ratio). These pathways regulate several cellular processes such as autophagy, metabolism, oxidative stress and gene expression that, in the end, are beneficial also for the improvement of cognitive and learning functions [190]. Considerable lines of evidence suggest that CR-associated benefits are mediated at the molecular level by the modulation of mTOR, which normally responds to variation of amino acid availability, oxygen, and growth signaling [79, 191]. Lower levels of mTOR activity, as would be expected to be encountered in CR, resulting in reduced protein and lipid synthesis, increased autophagy and stress- defense activity, enhanced regenerative capacity, and longer life in various organisms [192]. Several studies also indicate that mTOR influences feeding behavior and formation of memory in the hippocampus [193]. In Caenorhabditis elegans CR lifespan extension depended upon a group of regulators that are involved in stress responses and mTOR signaling, e.g., FOXO, Nrf1/2/3, FOXA, AMPK. CR reduced total oxygen consumption but increased the proportion of respiration devoted to ATP production. Apparently, CR drives the activity of key NAD⁺- associated mechanisms through a shift toward oxidative metabolism but does not necessarily increase overall respiration rates. In particular, CR-mediated extension in longevity was promoted by the up-regulation of FOXO, Nrf1/2/3, FOXA, AMPK, which are normally inhibited by mTOR signaling. The proposed importance of mTOR in CR also fits with the importance of the low- energy sensor AMPK [194], which inhibits mTOR in many species [79]. Hence, these findings support the view that CR extends lifespan in part through a reduction in mTOR signaling, and possibly insulin/IGF1 pathways [195].

Recent studies in mice also demonstrated that the reduction of insulin signaling by the action of CR shifts the energy metabolism in the brain from glucose to ketone bodies. CR additionally alleviates insulin resistance and through the inhibition of mTOR pathway, it enhances the clearance of misfolded protein including Aβ [196]. Therefore, CR was proposed to retard the progression of AD and is beneficial for healthy brain aging. Similar results were collected in ApoE-deficient mice (ApoE−/−), serving as a model of neurodegeneration [95]. ApoE−/− mice showed upon CR vs ad libitum feeding increased phosphorylation of AMPK and reduced activity of mTOR in the brain, which were associated with decreased Tau-phosphorylation [95]. Moreover, CR-mediated neuroprotective effects in ApoE−/− were characterized by increased numbers of PSD95-positive neurons and better cognitive performance [95].

Other than the above-cited effects mediated by CR, the role for mTOR in regulating healthy-aging and lifespan was highlighted by profiling genes differentially expressed in a successful model of aging. A genomic and transcriptomic data from 17 rodent species revealed a set of about 250 positively selected genes (PSGs) that are differentially regulated between long-lived naked mole-rats and short-lived rats [197]. In particular, long-lived rats show higher levels of these PSGs enriched for genes known to be related to aging. Among these enrichments were “cellular respiration” and “metal ion homeostasis”, as well as functional terms associated with processes regulated by the mTOR pathway: translation, autophagy and inflammation. Remarkably, among PSGs are RHEB, a regulator of mTOR, and IGF1, both central components of aging-relevant pathways [197]. Indeed, mTOR can be activated by RHEB either on the surface of the peroxisome [198]–in response to reactive oxygen species (ROS)–or on the surface of the lysosome [89]–in response to amino acids. Furthermore, among mTOR-regulated processes that are relevant for both growth and aging are translation and cellular respiration [79]. Consistent with the observed antagonistic expression patterns of PSGs in the long-lived naked mole-rat and short-lived rat, lower expression of genes related to these processes as well as pharmacological inhibition of the respective gene products were shown to be associated with longer lifespan [197].

Regarding human data, little research evaluated the potential role of diet on AD progression. There is evidence for certain dietary practices, such as Mediterranean diet and specific vitamin supplementation, offering protection against neurodegenerative disease development, but the effects of dietary intervention on the management of AD are relatively unknown [199, 200]. A few small clinical studies reportedly have shown a relationship between the ketogenic diet and improved cognition in AD patients [201]. These findings seem to be supported by studies showing AD patients consistently exhibit reductions in cerebral glucose utilization without alteration in brain ketone metabolism [202]. The application of CR to human studies has begun. To an extent, chronic CR has shown to be beneficial to human health such that moderate CR without malnutrition causes a protective effect against obesity, type-II diabetes, inflammation, hypertension and cardiovascular disease, all of which are major causes of morbidity, disability and mortality [203]. However, there remain valid concerns against CR since the duration and severity required for optimal health and anti-aging benefits is not feasible for most people over long periods of time and could lead to undesirable side effects [204].

5. Redox regulation of protein quality control networks in AD

mTOR signaling is strongly associated with neurodegeneration through its controlling protein homeostasis, which holds an important role in maintenance of brain structure and functions [101, 137, 205–207]. Protein homeostasis is preserved by a network of cellular apparati that control expression, localization, folding and interactions of proteins from their synthesis until their degradation [103, 208, 209]. The UPR operates to ensure the proper folding of proteins during the life of neurons [104, 210, 211], but if refolding strategies fail, misfolded/unfolded proteins are shifted to the UPS and autophagic degradation pathways [104, 105, 212–214]. AD is characterized by abnormal accumulation of aggregated proteins in postmitotic neurons, and the alteration of protein synthesis, folding, surveillance and degradation are implicated in this phenomenon [60, 105, 210, 215, 216]. The induction and functionality of protein quality control systems, and the preservation of proteostasis is closely associated with the redox balance of the brain [217, 218]. Low amounts of ROS can activate the adaptive cellular machinery to increase the organism’s stress resistance. This response involves the enhancement of antioxidant strategies and the induction of UPR, UPS and autophagy. Conversely, a chronic increase of ROS can on one side overwhelm the responsive capacity of these pathways and on the other side induce the oxidation of specific component of UPR, UPS or autophagy, thus exacerbating the accumulation of unfolded/misfolded proteins [60, 154] with deleterious consequences for cells, including neurons (Figure 1).

Figure 1.

In physiological conditions (depicted in the upper part of the figure), several oxidant stimuli, such as low levels of free radicals (ROS) and low amount of amyloid beta peptide (Aß), lead to coordinated stress responses (UPR, UPS and autophagy) for the removal of damaged molecules/organelles. In pathological conditions (depicted in the lower part of the figure), the overproduction of oxidants overwhelms the protein quality control, in part due to the oxidation/inactivation of some of its members, resulting in the accumulation of oxidized/dysfunctional proteins that are harmful for neuronal survival. In this scenario, mTOR hyperactivation, as it occurs in AD brain, not only contributes to impair autophagy machinery, but also dysregulates insulin signaling (by feedback mechanisms).

5.1. mTOR/autophagy

The crosstalk among OS and the mTOR/autophagy axis in AD remains understated, even though several studies highlighted the relevance of their association either in physiological or pathological conditions [60, 73, 102, 219]. It is well recognized that ROS induce autophagy and that autophagy, in turn, operates to reduce oxidative damage [217, 218, 220]. In particular, SH-SY5Y neuroblastoma cells exposed to OS accumulated more Aβ-enriched lysosomes than the control group, indicating the regulation of autophagy through OS [221]. Other authors have suggested an OS-induced improvement of APP processing and a decrease in Aβ elimination because of either reduced lysosomal degradation or Aβ structural modification [222]. Acute OS increase stimulates the initiation of autophagy, in association with stress signal pathways, through the posttranslational modifications of regulatory proteins [217]. Studies in transgenic mouse models of AD and AD-like dementia demonstrated that, in turn, the rescue of autophagic deficits led to a reduction of OS and brain deposition of Aβ while learning and memory deficits were prevented [118, 119, 223]. The redox proteomic analysis of AD brain samples demonstrates that the oxidative modification of key components of the protein quality control (PQC) pathways, as result of Aβ-induced increase of OS, might contribute to aberrant protein homeostasis and consequently to the deposition of senile plaques and NFTs [45, 103, 134, 154, 155, 224]. With respect to mTOR/autophagy, redox proteomics data suggest that increased OS, through the oxidative modification of autophagy components, might represent the link between the accumulation of Aβ, the alteration of mTOR signaling pathways and vice versa [224–226]. Indeed, the increase of oxidative stress appears to affect the final step of autophagic flux, which is involved in the degradation of the autophagosome cargo by lysosomal hydrolases. This final degradative process relies on the activity of two proteins: the vacuolar [H+] ATPase (V₀-ATPase) and cathepsin D, which were found oxidized in the brain of DS individuals at risk of developing AD, by redox proteomics approach [224–227]. In addition, V₀-ATPase is necessary for amino acids to activate mTOR, thus supporting that V₀-ATPase is an indirect, but integral, component of the mTOR pathway [228].

5.2. UPR

Accumulating evidence supports the concept of dysregulated UPR as key mechanism of neurodegeneration [213, 229]. Indeed, under persistent activation the UPR could foster the long-term memory and synaptic plasticity defects in AD [208, 210]. UPR is initiated upon disturbances of ER homeostasis and involves the activation of three transmembrane proteins in the ER membrane that serve as sensor proteins: the protein kinase-like endoplasmic reticulum kinase (PERK), the activating transcription factor 6 (ATF6), and the inositol-requiring kinase 1 (IRE1) [43, 230]. IRE1, PERK, and ATF6 pathways coordinate a complex network of stress signals to the cytoplasm and the nucleus that result in the inhibition of protein translation and control the expression of specific transcription factors to regulates folding and degradation of proteins. In this fashion, the UPR is finely linked to the proteolytic pathways, including the UPS and autophagy [231–235]. All three arms of the UPR are activated after BiP/Grp78 dissociation to allow the folding of accumulated unfolded proteins in the ER lumen. Despite its protective role, prolonged UPR induction is considered to promote the progression of neuropathological mechanisms [236]. Specific markers of UPR activation are increased in AD brain tissue. BiP/Grp78 expression levels are increased in the hippocampus and temporal cortex of AD and several studies have shown the increased presence of phosphorylated PERK, IRE1, and eIF2α in AD neurons [208, 210, 211, 237–239]. These observations imply that the UPR play a pathological role in the development of AD, since early stages. The adaptive response induced by UPR can modulate ROS production within the ER by reducing the folding demand and upregulating the expression of antioxidant factors [236, 240]. The control of the redox balance by UPR is essentially linked to IRE1 and PERK pathways through the induction of Nrf2-related antioxidant response and ATF4, which plays a key role in glutathione (GSH) synthesis [208, 230, 239]. However, studies on AD brain highlighted that under a condition of chronic UPR activation, an uncoupling between PERK and Nrf2 signals occurs, while ATF4 switches from pro-survival to pro-apoptotic pathways [211]. These data support the inability of UPR to induce stress responses, favoring the build-up of pro-oxidant species and the damage of biological components. Consistent with these findings, redox proteomic studies reported increased HNE modification of BiP/Grp78 during AD pathology [224], which may cause the failure to bind to misfolded proteins thus contributing to the dysfunction of the UPR. In addition, the oxidation of eIF2α, observed in MCI IPL, might lead to its impaired activity leading to UPR defects and exacerbating the accumulation of unfolded/misfolded proteins [241]. Emerging studies also demonstrated a mechanistic link between the ER stress and autophagy in which the activation of PERK and its downstream targets induce autophagy [242] by regulating the gene transcription of autophagy components such as BECN1, MAP1LC3B, ATG5, ATG7 and ATG12. Therefore, autophagy is induced under conditions of ER stress and reports on neurons from human brain demonstrated that high autophagic flux is associated with activated UPR. Furthermore, the UPS overload observed during ER stress conditions in AD may trigger the induction of autophagy to degrade ubiquitinylated proteins [208, 210, 234]. Within this context, AD is characterized by the establishment of a vicious cycle that arises by the increased activation of the UPR combined with decreased efficiency of autophagy, as well as other parts of the proteostasis network, due to the aging process and the increased OS [60, 209, 230].

5.3. UPS

UPS represents one of the main pathways of protein clearance in eukaryotic cells, which not only digests misfolded, oxidized, or damaged proteins, but also eliminates proteins involved in a plethora of intracellular processes [104]. Experimental evidence indicates that disturbance of UPS plays a major role in AD [57, 58, 212, 216, 231, 232]. The activity of proteasome is significantly decreased in the brain of AD [243], and it was proven that proteasome inhibition leads to the accumulation of Aβ and tau. Increased OS/NS levels were suggested as one of the major factors contributing to the impairment of UPS in AD [244]. Indeed, if from one side moderately oxidized proteins were preferentially recognized and degraded by the proteasome, severely oxidized proteins cannot be easily degraded and, instead, inhibit the proteasome [245, 246]. Therefore, the overload of unfolded/oxidized substrates coupled with the oxidative damage of UPS constituents may lead to accumulation of abnormal proteins and to selective degeneration of neurons. The proficiency of the UPS also depends on the activity of deubiquitinating enzymes that play an essential role in determining the rate of protein clearance in cells [58]. Among deubiquitinating enzymes, ubiquitin carboxy-terminal hydrolase L1 was found oxidatively modified in AD brain, and as a consequence its activity was found to be reduced [57, 155, 224, 225, 247]. In addition, data obtained by Saito et al. showed that the E3 ubiquitin ligase—HRD1—was precipitated from solution by OS, [248]. Cecarini et al. [249] demonstrated a significant reduction in proteasome-mediated degradation of oxidized proteins in MCI and AD subjects, and such decreased proteolytic activities were associated with the increase of oxidative modifications such as carbonylation, HNE-modification, and neuroprostane conjugation, thereby confirming a role for OS as a causative factor.

Despite that UPS and autophagy were often considered as two independent proteolytic pathways, recent advances strongly suggest that their activities are carefully orchestrated by multiple crosstalk mechanisms [214]. Indeed, both the pathways share common substrates such as ubiquitinylated proteins and, under specific conditions, they both can selectively degrade short-lived proteins and/or long-lived proteins [104]. In neurodegenerative conditions, where the accumulation of toxic species is profound, cells can reorganize proteolysis regulating the communication between the two protein degrading pathways [90, 159, 210, 213, 214, 219]. In principle, when one proteolytic system is damaged and shows a reduced functionality, the enhanced activity of the other pathway may become a compensatory mechanism necessary to protect neuronal cells against the accumulation of toxic species [231, 250]. An example of the inter-regulation between the UPS and autophagy is the observation that impairment in UPS-mediated degradation leads to an increased autophagic function. The activation of this compensatory mechanism involving autophagy activation allows cells to reduce the number of aggregates formed in response to proteasomal inhibition [104, 124, 208]. The analysis of SH-SY5Y cells overexpressing either the wild-type AβPP gene or the 717 valine-to-glycine AβPP-mutated gene demonstrated that, in addition to the increase of oxidative stress, neuronal cells demonstrated a reorganization of the cellular proteolytic machineries with marked inhibition of proteasome activities, impairment in the autophagic flux and increased expression of compensatory mechanisms as an attempt to rescue proteolysis [251]. The analysis of AD post-mortem brain suggests that the accumulation of amyloid-β plaques arises by the concomitant impairment of both the UPS system and autophagy and a prominent role in this cascade of events might be played by the increased oxidative damage of protein degradation machineries components [57, 74, 91, 92, 103, 161, 216, 232, 252].

6. Conclusions

Several studies have demonstrated the complex etiology of Alzheimer disease as the result of multifaceted toxic mechanisms. This likely explains the failure of several therapies that if correctly target a particular pathological pathway, still leaves ongoing pathological events that act independently from the “selected target”. In this scenario, it becomes a future challenge to design therapeutic strategies that have the potential to intersect multiple pathways if we are able to identify the correct “hub”. Collecting studies suggest that a major cellular hub is represented by OS and mTOR signaling. This signaling network is complex, with several downstream physiological outputs, and the mechanisms underlying its age-related effects still remain enigmatic. These pathways are intimately linked within a complex network of signals that are essential for a healthy longevity. It is likely that alteration of one or more of the components of the mTOR pathway is able to modulate the entire system.

Growing evidence suggest that hyperactivation mTOR signaling pathway play a key role in major pathological mechanisms underlying AD. This aberrant activation correlates with dysfunction of energy metabolism, reduced ATP production and increased oxidative stress levels in a sort of self-sustaining cycle. Understanding the precise molecular mechanisms that foster this vicious cycle is essential to develop mTOR-targeted therapies not only to treat Alzheimer disease but also other neurodegenerative disorders.

HIGHLIGHTS:

• Oxidative stress in brain is critically involved in Alzheimer disease progression.

• mTORC1 activation in AD & MCI brains leads to inhibition of protein quality control

• mTORC1 and mitochondrial functions have significant crosstalk, including metabolism

• mTORC1 activation is related to brain insulin resistance and neuronal death in AD.

• Selective mTORC1 inhibition activation may be a promising AD therapeutic strategy.