Tiny Messengers, Huge Consequences: Extracellular Vesicles and mTOR Signaling in Neuroinflammation

ABSTRACT

Neuroinflammation plays a fundamental role in the pathogenesis of neurodegenerative disorders. Among the central regulators of this process are the mechanistic target of rapamycin (mTOR) signaling pathway and extracellular vesicles (EVs). This review discusses the bidirectional interaction between mTOR and EVs, describing their interconnection in the regulation of cellular communication and inflammatory responses in the central nervous system (CNS). On one hand, mTOR controls EV biogenesis and cargo composition, and, on the other hand, EVs also modulate mTOR activity in target cells with effects in neuronal survival, glial activation, and immune signaling. This bidirectional communication creates a feedback loop that, depending on cell context and molecular cues, may either promote neuroprotection or exacerbate neurotoxicity and inflammation.

Neuroinflammation is a crucial mediator in the pathogenesis of multiple neurodegenerative diseases, acting not only as a consequence but also as a driver of disease progression. Yet, much remains to be understood about the molecular and cellular events involved. The goal of this review is to explore the dynamic crosstalk between mechanistic target of rapamycin (mTOR) signaling and extracellular vesicles (EVs) in shaping the outcomes of inflammatory processes within the pathological central nervous system (The image of the graphical abstract was done with BioRender.).

Abbreviations

4E‐BP1

4E‐binding protein 1

6‐OHDA

6‐hydroxydopamine

ACDVs

artificial cell‐derived vesicles

ADEVs

astrocyte‐derived EVs

Akt

protein kinase B

AlCl3

aluminum chloride

ALIX

ALG‐2‐interacting protein X

ALS

amyotrophic lateral sclerosis

Aβ

amyloid beta

BBB

blood–brain barrier

BDNF

brain‐derived neurotrophic factor

CNS

central nervous system

CSF

cerebrospinal fluid

DNA

desoxyribonucleic acid

ELISA

enzyme‐linked immunosorbent assay

EPs

extracellular particles

ESCRT

endosomal sorting complex required for transport

EVs

extracellular vesicles

GFAP

glial fibrillary protein

HD

Huntington's disease

HSC70

heat shock 70 kDa protein

HSP90β

heat shock protein 90β

IBA1

ionized calcium‐binding adapter molecule 1

IL‐1β

interleukin‐1 beta

IL‐6

interleukin‐6

ILVs

intraluminal vesicles

ISEV

International Society for Extracellular Vesicles

L1CAM

L1 cell adhesion molecule

LAMP2

lysosome‐associated membrane protein 2

LRRK2

leucine‐rich repeat kinase 2

MALAT1

metastasis‐associated lung adenocarcinoma transcript‐1

MDEVs

microglia‐derived EVs

miRNAs

microRNAs

MISEV2023

minimal information for studies of extracellular vesicles 2023

mRNAs

messenger ribonucleic acids

MSC‐EVs

mesenchymal stem cell‐derived extracellular vesicles

mTOR

mechanistic target of rapamycin

mTORC1

mTOR complex 1

mTORC2

mTOR complex 2

NC

non‐coding RNAs

NDEVs

neuron‐derived EVs

NVEPs

non‐vesicular extracellular particles

ODEVs

oligodendrocyte‐derived EVs

PBMCs

peripheral blood mononuclear cells

PD

Parkinson's disease

PELI1

pellino E3 ubiquitin protein ligase 1

PI3K

phosphatidylinositol 3‐kinase

POMC

proopiomelanocortin

p‐tau

phosphorylated tau

Rab

Ras‐related in brain

S1P

sphingosine 1‐phosphate

S6K

ribosomal protein S6 kinase

S6K1

S6 kinase 1

SIMOA

single‐molecule array

SNARE

soluble NSF attachment protein receptor

SVs

synthetic vesicles

TDP43

TAR DNA‐binding protein 43

TLRs

toll‐like receptors

TMEM119

transmembrane protein 119

TNF‐α

tumor necrosis factor‐α

TREM2

triggering receptor expressed on myeloid cells 2

TrkB

tropomyosin receptor kinase B

TSG101

tumor susceptibility gene 101 protein

1. Introduction

mTOR signaling stands out as one of the central regulatory pathways within the cell, governing both physiological and pathological conditions. Among the latter, neuroinflammation is a common hallmark of different neurodegenerative conditions such as Alzheimer's disease (AD), Parkinson's disease (PD), Huntington's disease (HD), and multiple sclerosis (MS). Indeed, mTOR signaling regulates neuroinflammatory processes in these conditions since it contributes to the production of inflammatory mediators.

Extracellular vesicles (EVs) are membranous vesicles released by all types of cells, including neurons and glial cells. Indeed, in the central nervous system (CNS), they have been proven to be key players in controlling cell‐to‐cell communication. Apart from their physiological role, which has been reviewed elsewhere (Lizarraga‐Valderrama and Sheridan ²⁰²¹), EVs are increasingly being considered as drivers of pathological conditions, particularly neuroinflammation. In this line, cellular stress affects the sorting of specific proteins and other components within EVs, directly defining the biological response in recipient cells (Frühbeis et al. 2013; Malenka and Bear ²⁰⁰⁴; Eldh et al. ²⁰¹⁰).

Regarding mTOR, it directly controls key events in EV biogenesis and cargo sorting. Excessive mTOR activity suppresses multivesicular body (MVB) fusion with lysosomes and hence blocks EV secretion, whereas inhibition of mTOR enhances EV release in neurons (Solana‐Balaguer, Campoy‐Campos, et al. ²⁰²³; Solana‐Balaguer, Martín‐Flores, et al. ²⁰²³; Zou et al. ²⁰¹⁹). However, in a study conducted by Leng et al. (2024) they discovered that in iPSC‐derived astrocytes, mTOR hyperactivation suppresses autophagic flux, which presumably reroutes endolysosomal contents to MVBs and increases EV secretion. This highlights that mTOR activity dynamics influence EV release in a cell type–specific manner. Moreover, mTOR activity controls EV cargo content, influencing the selective enrichment of proteins, ribonucleic acids (RNAs), and signaling molecules.

Pathologically, mTOR signaling dysregulation may facilitate the release of EVs containing pro‐inflammatory content, which may promote neuroinflammation and contribute to neurotoxicity.

Therefore, a fine‐tuned reciprocal influence between EVs and mTOR signaling does exist, ultimately setting the balance between neurodegeneration and neuroprotection.

This review, therefore, aims to examine the interplay between mTOR signaling and the intercellular communication mediated by EVs in the context of neuroinflammation and neurodegeneration.

2. Extracellular Vesicles: Biogenesis, Types, and Functions

EVs are heterogeneous lipid bilayer membrane‐enclosed vesicles released in an evolutionarily conserved manner by all cell types (Yáñez‐Mó et al. ²⁰¹⁵). EVs were first believed to be vehicles for waste elimination produced by the same cells (Johnstone et al. 1987). However, EVs have been revealed as important modulators of physiological processes such as synaptic plasticity, myelination, and neurogenesis (Regev‐Rudzki et al. ²⁰¹³; Cicero et al. ²⁰¹⁵; Colombo et al. ²⁰¹⁴).

2.1. EV Classification: Exosomes, Microvesicles, Apoptotic Bodies

Recent advances in isolation and analytical methods have allowed the identification of many types of vesicles which can be classified by size, cellular origin, biosynthesis, and composition. With the aim of finding consensus and reproducibility in EVs studies, the International Society for Extracellular Vesicles (ISEV) has made tremendous work putting together the minimal recommendations in its “Minimal information for studies of extracellular vesicles” MISEV2023 guide (Welsh et al. 2024).

Based on these recommendations, the MISEV2023 guidelines advocate for the use of the term “EVs” to describe a population of “particles that are released from cells, are delimited by a lipid bilayer, and cannot replicate on their own”. A broader term, extracellular particles (EPs), encompasses all particles released from cells, regardless of whether they are enclosed by a lipid bilayer. Within this classification, EPs include both EVs and non‐vesicular extracellular particles (NVEPs), the latter lacking a lipid bilayer. In addition, among lipid‐bilayer vesicles, EV mimetics—defined as EV‐like particles produced through direct artificial manipulation—can be further categorized into artificial cell‐derived vesicles (ACDVs), generated under conditions of induced cell disruption (e.g., extrusion), and synthetic vesicles (SVs), which are synthesized de novo from molecular components or designed as hybrid entities. Moreover, ISEV recognizes the use of “operational terms” as a prefix to “EVs” when clearly distinct EV subtypes are separated based on characteristics such as size, density, molecular composition, or cellular origin. Nonetheless, these “operational terms” should be applied with caution. For example, the MISEV2023 guidelines acknowledge the extensive body of work employing terms such as “small EVs” or “large EVs,” but they also note that many separation methods yield EV populations with overlapping size distributions, and that there is no clear consensus on the upper and lower size cutoffs for these categories. In general, however, the guidelines state that small EVs may refer to vesicles with a diameter of 200 nm or less, thereby classifying large EVs as those larger than 200 nm. Finally, concerning the classification of EVs based on their biogenesis pathways, it is recommended to rely on robust evidence. This is due to the fact that most EV separation techniques fail to enrich for EVs generated by distinct mechanisms. Furthermore, the definitive characterization of biogenesis‐based subtypes poses challenges, as there are no universal molecular markers for ectosomes, exosomes, or other EV subtypes. Yet, given their distinct subcellular origins, a brief description of the principal EV subtypes will be provided in this and the following sections.

Exosomes can be defined by their size, ranging from 30–80 (small exosomes) to 90–120 nm (large exosomes) (Yáñez‐Mó et al. ²⁰¹⁵). They are produced by the endosomal system, generated from the inward budding of the limited MVB membrane forming intraluminal vesicles (ILVs). These ILVs will later fuse with the plasma membrane and be released into the extracellular space (Zhang et al. 2019). More recently, Kalluri and LeBleu (2020) have proposed a distinct classification of exosomes based on their size in ExoA (40–75 nm), ExoB (75–100 nm), and ExoC (100–160 nm). Moreover, they have also proposed a classification based on content, functional, and source heterogeneity.

Ectosomes result from outward budding from the plasma membrane released into the extracellular environment (Jeppesen et al. 2023). Depending on the size of the vesicle, ectosomes have been classically classified into microvesicles, small ectosomes (100–1000 nm) rich in lipid raft domains (Del Conde et al. 2005), and large oncosomes, 1–10 μm non‐apoptotic membrane blebs derived exclusively from tumor cells capable of propagating tumor‐promoting material (Al‐Nedawi et al. ²⁰⁰⁸; Minciacchi et al. ²⁰¹⁵). Nonetheless, they are more diverse in size and function than what was initially believed (Meldolesi 2018). Various subpopulations of ectosomes have been identified, such as migrasomes (Ma et al. 2015), exopheres (Melentijevic et al. 2017), and midbody remnants (Rai et al. 2021). For instance, migrasomes are vesicles with a diameter up to 3 μm released from the retraction fiber located in the rear of migrating cells. These vesicles contain several smaller vesicles and cytosolic content inside which are accumulated during the cell migration process (Zhang et al. 2023). On the other hand, exopheres are 3.5–4 μm in diameter and are released in response to neurotoxic and metabolic stress. These vesicles are used to eliminate neurotoxic components when proteostasis and organelle function are compromised (Melentijevic et al. 2017).

Apoptotic bodies are 0.5–5 μm vesicles formed by the outward bending of the membrane from cells undergoing apoptosis. Several studies have reported that apoptotic bodies are involved in the inhibition of inflammatory responses and the transfer of histones, DNA, and cytosolic proteins to phagocytes (Wen et al. 2023).

2.2. EVs Biogenesis

As mentioned above, exosomes and ectosomes have different biogenesis processes. Exosomes are generated in the endosomal pathway as ILVs and released when MVB fuses with the cell membrane, while microvesicles are produced through outward budding from the plasma membrane (Raposo and Stoorvogel 2013).

The biogenesis of exosomes can be executed via two pathways: the endosomal sorting complex required for transport (ESCRT)‐dependent and the ESCRT‐independent pathway (van Niel et al. ²⁰¹⁸). The ESCRT‐dependent pathway involves the recruitment of ESCRT complexes to the endosomal membrane to form the ILVs. On the contrary, in the ESCRT‐independent pathway, the formation of ILVs is mediated by tetraspanin‐enriched microdomains (van Niel et al. ²⁰¹¹) and lipid rafts (Takeda et al. 2008). Upon the formation of ILVs, MVB can be later fused with lysosomes for degradation or merge with the plasma membrane to release these ILVs as exosomes (Han et al. 2022). Indeed, exosome formation is a complex and dynamic process that involves both canonical ESCRT‐dependent mechanisms and alternative, ESCRT‐independent pathways. Furthermore, these pathways are not mutually exclusive; rather, they coexist and interact dynamically, with the balance between them being modulated by cellular context, cargo type, and external stimuli (see Arya et al. ²⁰²⁴ for a review). However, it is also worth mentioning that the primary synthesis pathway appears to be the ESCRT‐dependent machinery, with the independent pathways acting to modulate or refine the process under specific cellular contexts rather than serving as the basal mode of exosome secretion (Gurung et al. 2021).

The secretion of exosomes to the extracellular space is mediated by the Rab GTPases, which are essential for intracellular vesicle transport (Blanc and Vidal 2018) and by soluble NSF attachment protein receptor (SNARE) protein, indispensable for the fusion of MVBs with the plasma membrane (Xie et al. 2022).

Once EVs are in the extracellular space, they bind to their target cells via specific surface receptors, triggering various intracellular signaling pathways (Liu and Wang 2023). Afterwards, EVs are taken up by direct fusion with the plasma membrane or endocytosed by the cell.

To date, endocytosis is the most common mechanism in which EVs can be internalized by other cells (Joshi et al. 2020). Endocytic uptake of EVs can occur through different mechanisms, including clathrin‐mediated endocytosis, lipid‐raft mediated endocytosis, caveolin‐mediated endocytosis, phagocytosis, or micropinocytosis. These different mechanisms of EVs entry are not always mutually exclusive and can co‐exist (Gurung et al. 2021). It is worth noting that we will emphasize the role of mTOR signaling in exosomes, given its involvement in their biogenesis, as will be discussed in subsequent sections. Still, in accordance with the MISEV2023 guidelines, the term “exosomes” will only be used when their endosomal origin can be unequivocally established.

2.3. Molecular Cargo of EVs

EVs typically contain lipids, nucleic acids, and proteins derived from the EV cellular origin. However, their composition can vary depending on the cell type, the mechanisms of secretion, or even EVs' localization (Zhang et al. 2019).

In the case of exosomes, they originate from MVB, which is mainly regulated by ESCRT proteins, although ESCRT‐independent pathways do exist. These proteins and their accessory proteins (ALG‐2‐interacting protein X [ALIX], tumor susceptibility gene 101 protein [TSG101], heat shock 70 kDa protein [HSC70], and heat shock protein 90β [HSP90β]) are found in all EVs regardless of the type of cell from which they originate (van Niel et al. ²⁰⁰⁶). Moreover, exosomes are highly enriched in tetraspanins (CD9, CD63, CD81, CD82), which are involved in endosomal vesicle trafficking and cell internalization, as well as proteins responsible for membrane docking and fusion (RABs, annexins) (Vlassov et al. 2012). Other exosomal proteins include metabolic enzymes, ribosomal proteins, transmembrane, adhesion, signal transduction, and cytoskeletal molecules (Mathivanan et al. 2010). Additionally, exosomes and other EVs can carry different patterns of nucleic acids such as DNA, messenger ribonucleic acids (mRNAs), and non‐coding RNAs (ncRNAs) (Mashouri et al. 2019). In fact, microRNAs (miRNAs), a class of ncRNAs, are the most abundant RNA species in exosomes (Huang et al. 2013).

2.4. Novel EVs Functions Beyond Their Tissue of Origin: The Example of Brain‐Derived EVs as Biomarkers in Blood

One of the most promising aims in the field of EVs is to isolate brain cell‐specific EVs in peripheral fluids like blood to use them as readouts of health and disease, as non‐invasive reliable biomarkers to monitor disease progression (Fiandaca et al. 2015). However, this is a challenging goal due to the presence of particles with a similar EVs profile, such as non‐vesicular extracellular particles, EVs released by platelets, and the presence of lipoproteins, with a similar EVs profile, (reviewed in Manolopoulos et al. ²⁰²⁵; Kumar et al. ²⁰²⁴) and the most important fact, the lack of consensus for their isolation techniques (Badhwar et al. 2024). Researchers are currently using antibodies against the extravesicular domains of transmembrane proteins to immunoprecipitate them and proceed to analysis. Hence, using antibodies against L1 cell adhesion molecule (L1CAM) for neuron‐derived EVs (NDEVs) (Shi et al. 2014; Nogueras‐Ortiz et al. ²⁰²⁴; Fiandaca et al. ²⁰¹⁵), or glial fibrillary protein (GFAP) for astrocyte‐derived EVs (ADEVs) (Delgado‐peraza et al. ²⁰²¹; Willis et al. ²⁰¹⁷; Vos et al. ²⁰⁰⁴) is becoming quite established. However, many questions arise from this strategy that need to be solved. For example, the topology of these transmembrane proteins, since the intracellular and extracellular parts of several receptors can switch through the process of endocytosis and vesicular production (Cvjetkovic et al. 2016). In this line, NDEVs can reflect neuronal health or carry disease‐specific proteins like Amyloid beta (Aβ), tau, α‐synuclein (α‐syn), and TAR DNA‐binding protein 43 (TDP43), varying across conditions such as AD, PD, and amyotrophic lateral sclerosis (ALS). ADEVs can also contain pathological proteins and are isolated using astrocyte‐specific markers such as GFAP (Delgado‐peraza et al. ²⁰²¹; Willis et al. ²⁰¹⁷; Vos et al. ²⁰⁰⁴). Microglia‐derived EVs (MDEVs) provide insight into neuroinflammation and neuroprotection but face challenges in detection and specificity due to overlapping with peripheral immune cells (Ghosh and Pearse 2024). Certainly, the fact that immune‐related markers are not exclusive to microglia, and peripheral macrophages exhibit similar profiles, evidences the necessity to use a combination of markers (e.g., TMEM119, IBA1) and functional assays to attribute the EVs to a microglial origin unambiguously (Ghosh and Pearse 2024). Functionally, MDEVs have been shown to carry increased levels of cytokines such as interleukin‐1 beta (IL‐1β), tumor necrosis factor‐α (TNF‐α), and inflammatory miRNAs including miR‐155 and miR‐146a‐5p, which can amplify inflammatory cascades and disrupt normal neuronal signaling (Brites and Fernandes 2015). Oligodendrocyte‐derived EVs (ODEVs) may indicate myelin integrity and are relevant to diseases like multiple sclerosis and traumatic brain injury (Casella et al. 2020). Particularly, ODEVs derived from multiple sclerosis patients' blood contain increased levels of myelin basic protein (Agliardi et al. 2023). Despite their potential, further validation is needed for clinical application.

Another issue is the actual specificity of these markers, since some tumoral cells can express some of them, such as L1CAM (Fogel et al. 2003; van der Maten et al. ²⁰¹⁹). And finally, another drawback is the limited proportion of these EVs in the serum or plasma among the whole circulating EVome secreted by each cell of the body. Another question that arises is whether neurodegeneration not only is able to change the EVs content from the brain but also the content of the EVs from peripheral cells such as peripheral blood mononuclear cells (PBMCs), for example, due to neuroinflammation and the associated immune response.

Therefore, the advantage of isolating brain‐derived blood circulating EVs as biomarkers is that their isolation is much less invasive in comparison to a lumbar puncture to collect cerebrospinal fluid (CSF). Moreover, they could be used for early detection in prodromal phases of neurodegeneration, before clinical symptoms appear. Hence, their future use is promising to monitor treatment efficiency and disease progression, although it will require high‐sensitivity methods, like enzyme‐linked immunosorbent assay (ELISA), mass spectrometry, or single‐molecule array (SIMOA) detection methods.

Hence, brain‐derived EVs isolated from blood offer a minimally invasive window into brain health, with significant potential for diagnosing, prognosing, and monitoring neurological diseases. Yet, further validation and standardization are needed for clinical translation. Whether these circulating brain‐derived EVs have an effect in recipient peripheral cells or whether they can modulate microenvironments for immune responses that could contribute to cope or accelerate neurodegeneration is still unknown.

3. The mTOR Signaling Pathway in the Regulation of EVs Cargo and Neuroinflammation

mTOR has emerged as a pivotal signaling hub that orchestrates a wide array of cellular functions crucial to CNS homeostasis, including protein synthesis, autophagy modulation, metabolism, and cell survival. In the CNS, proper regulation of mTOR activity is essential for maintaining neuronal integrity and synaptic plasticity, whereas aberrant mTOR signaling is increasingly implicated in the pathogenesis of neuroinflammation and neurodegeneration (Hodges and Lugo 2020).

3.1. Overview of mTOR Signaling in the CNS

The mechanistic target of rapamycin (mTOR) is a serine/threonine kinase that functions as the catalytic core of two distinct multiprotein complexes: mTOR complex 1 (mTORC1) and mTOR complex 2 (mTORC2) (Bockaert and Marin 2015). mTORC1, characterized by the presence of Raptor, among other interactors, is highly sensitive to fluctuations in nutrient levels and growth signals, and its activation drives protein synthesis by phosphorylating downstream effectors such as S6 kinase 1 (S6K1) and 4E‐binding protein 1 (4E‐BP1) while concurrently suppressing autophagy in nutrient‐rich conditions (Bockaert and Marin 2015). In contrast, the mTORC2 complex, which includes Rictor and mSIN1, is more dedicated to promoting cell survival and regulating cytoskeletal dynamics, thereby ensuring the structural integrity and adaptive function of neurons (Bockaert and Marin 2015). Through the integration of extracellular cues such as growth factors, amino acids, oxidative stress, and inflammatory cytokines, mTOR serves as a master regulator that coordinates anabolic and catabolic cellular processes critical for neuronal growth, synaptic plasticity, microglial priming, and overall CNS functionality (Bockaert and Marin 2015). The precise balancing of these processes is vital for normal neuronal operation, as dysregulation can lead to aberrant protein synthesis or autophagic dysfunction, both of which have been linked to disease states (Bockaert and Marin 2015).

3.2. mTOR in Neuroinflammation

Neuroinflammation is a hallmark of many CNS disorders and is typically characterized by the overactivation of glial cells—especially microglia and astrocytes—that release pro‐inflammatory cytokines detrimental to neuronal health (Zeng et al. 2022). When mTOR signaling becomes hyperactivated, as observed in conditions such as epilepsy, traumatic brain injury, and some neurodegenerative disorders, it not only contributes to aberrant protein synthesis and reduced autophagy but also amplifies the production of inflammatory mediators (Palavra et al. 2016). For example, chronic neuroinflammation, driven by sustained mTOR overactivation, contributes to a toxic environment that accelerates neuronal degeneration and synaptic dysfunction (Zeng et al. 2022). In addition, hyperactive mTOR signaling is implicated in compromising the integrity of the blood–brain barrier (BBB), facilitating the infiltration of peripheral immune cells and thereby intensifying the neuroinflammatory milieu (Zeng et al. 2022).

The pathogenic role of mTOR in neuroinflammation is further underscored by its complex crosstalk with other signaling cascades, notably the nuclear factor kappa‐light‐chain‐enhancer of activated B cells (NF‐κB) pathway and the phosphatidylinositol 3‐kinase (PI3K)/protein kinase B (Akt) axis (Zeng et al. 2022). Under conditions of cellular stress, mTOR activation can potentiate NF‐κB signaling, a key driver of pro‐inflammatory gene transcription, thereby enhancing the synthesis of inflammatory cytokines such as interleukin‐6 (IL‐6), TNF‐α, and IL‐1β, which amplify the inflammatory response in the CNS (Hodges and Lugo 2020). Additionally, the PI3K/Akt pathway functions upstream to phosphorylate and activate mTOR, creating a positive feedback loop that further intensifies inflammation while also influencing cell survival and metabolic homeostasis (Zeng et al. 2022). This bidirectional interaction means that inflammatory stimuli often lead to mTOR hyperactivation, and in turn, activated mTOR promotes further inflammatory mediator production; this reciprocal relationship establishes mTOR as a central nexus in neuroinflammatory signaling that critically shapes the course of CNS disorders (Hodges and Lugo 2020).

Despite the above protective aspects associated with mTOR inactivation, emerging evidence indicates that complete or prolonged inactivation of mTOR may paradoxically promote neuroinflammatory conditions. This paradox arises because mTOR signaling is intricately involved not only in promoting pro‐inflammatory responses but also in mediating cell survival, lysosomal function, and immune homeostasis, particularly in microglia (Shi et al. 2022). For instance, in PD, inactivation of some of the mTOR activities with rapamycin, which mostly inhibits mTORC1 activities (Malagelada et al. 2010; Yang et al. ²⁰²¹), protects against dopaminergic neurotoxin‐induced neuron death both in vitro and in vivo (Malagelada et al. 2010). In the same line, in the context of AD, mTOR inactivation has been observed to impair microglial phagocytic clearance of Aβ plaques by reducing lysosomal biogenesis and the expression of receptors like triggering receptor expressed on myeloid cells 2 (TREM2), thereby exacerbating plaque accumulation and neuroinflammation (Shi et al. 2022).

Further complexity is introduced by the fact that mTOR inactivation does not exert uniform effects across all cell types in the brain. In neurons, for example, acute mTOR inactivation can promote autophagy and facilitate the clearance of misfolded proteins; however, in microglia, the same inactivation may impair lysosomal function and reduce phagocytic capacity, leading to the accumulation of neurotoxic debris (Shi et al. 2022). Moreover, the temporal control of mTOR inhibition is also a crucial aspect. Indeed, while acute, partial mTOR inhibition is neuroprotective, prolonged mTOR inhibition is a double‐edged sword. Specifically, regarding the role of acute mTOR inactivation in neuronal death, it can confer neuroprotection via the activation of autophagy and the preservation of the bioenergetic status, especially in contexts of metabolic and excitotoxic stress (Zheng et al. 2016; Hwang et al. ²⁰¹⁷). Furthermore, we have shown that acute treatment with rapamycin is enough to block the protein synthesis of some pro‐apoptotic proteins while preserving key survival signals such as Akt phosphorylation at Thr308 (Malagelada et al. 2010). Contrarily, extended blockade risks interference with essential anabolic processes that depend on mTOR signaling, including synaptic plasticity and long‐term protein synthesis. Moreover, evidence suggests that prolonged treatment with strong inhibitors such as Torin1 may lead to increased toxicity over time (Malagelada et al. 2010), meaning that while an early acute intervention can prevent the initiation of cell death signals, continuous prolonged inhibition might impair neuronal functions that are necessary for long‐term survival.

Therefore, the net effect of mTOR inhibition on neuroinflammation depends critically on the cellular context, the timing of inhibition, and the balance between autophagic flux, cytokine production, and the maintenance of immune surveillance mechanisms (Shi et al. 2022). These findings underscore the importance of fine‐tuning mTOR pathway modulation in therapeutic interventions, considering the cell‐specific roles of mTOR in the CNS.

3.3. mTOR as a Regulator of EVs Biogenesis and Cargo Sorting

mTOR signaling itself exerts significant regulatory influence over the biogenesis, release, and cargo composition of EVs (Zou et al. 2019). Under conditions of elevated mTORC1 activity, the fusion of MVBs with lysosomes is favored as opposed to their fusion with the plasma membrane, leading to a suppression of exosome secretion. Conversely, inhibition of mTORC1—such as by nutrient deprivation or pharmacological intervention—shifts MVB trafficking toward the plasma membrane, thereby enhancing the release of exosomes into the extracellular space (Solana‐Balaguer, Campoy‐Campos, et al. ²⁰²³; Solana‐Balaguer, Martín‐Flores, et al. ²⁰²³). Moreover, mTOR activity influences the selective packaging of specific proteins and RNAs into EVs, meaning that dysregulated mTOR signaling can lead to an altered EV cargo profile that might skew intercellular communication toward a pro‐inflammatory state (Zou et al. 2019). For example, when mTOR is hyperactivated, EVs may become enriched in pro‐inflammatory microRNAs and cytokines, which can then propagate inflammatory signals across neural networks and exacerbate neurotoxicity (see Marangon et al. ²⁰²² for a review). This regulatory role underscores the dual function of mTOR as both a target for EV‐delivered signals and as a critical determinant of the composition and functional properties of EVs.

3.4. Extracellular Vesicles as Modulators of mTOR Activity

EVs have garnered significant attention as potent mediators of intercellular communication within the CNS due to their ability to shuttle bioactive molecules between cells. In the context of neuroinflammation, EVs derived from activated microglia have been shown to carry specific microRNAs—most notably miR‐124‐3p—that can influence mTOR signaling in recipient neurons (Huang et al. 2018). The transfer of miR‐124‐3p via microglial EVs results in the downregulation of mTOR activity within target neurons, which in turn reduces the production of pro‐inflammatory cytokines and fosters neuroprotective processes such as enhanced neurite outgrowth and axonal regeneration (Huang et al. 2018). This miRNA‐mediated modulation demonstrates that EVs are not passive carriers; rather, they actively participate in regulating key intracellular signaling pathways that determine the inflammatory state and viability of neuronal cells. In this line, NDEVs also contain synaptic proteins and trophic factors that can contribute to neuronal trophic support when internalized in the soma, dendrites, or even in dendritic spines by the recipient neuron. NDEVs contain brain‐derived neurotrophic factor (BDNF) and its receptor tropomyosin receptor kinase B (TrkB), and BDNF–TrkB EV‐mediated signaling is involved in the potentiation of excitatory synaptic contacts and the formation of dendritic spines in cortical neurons (Solana‐Balaguer, Campoy‐Campos, et al. ²⁰²³; Solana‐Balaguer, Martín‐Flores, et al. ²⁰²³).

3.5. Reciprocal Regulation and Impact on Neurotoxicity and Neuroinflammation

The interplay between mTOR signaling and EV dynamics is thereby inherently bidirectional and exerts a profound influence on both neuroinflammation and neurotoxicity. On one level, EVs carrying regulatory molecules such as miR‐124‐3p can be internalized by neurons and glial cells, leading to the suppression of mTOR activity and a consequent reduction in inflammatory mediator release. This effect has been observed in models of traumatic brain injury, where the EV‐mediated delivery of miR‐124‐3p is associated with decreased neuroinflammation and enhanced neural repair (Huang et al. 2018). Furthermore, in vitro, hypothalamic proopiomelanocortin (POMC) neurons may transport and absorb adipocyte‐derived EVs. The lncRNA metastasis‐associated lung adenocarcinoma transcript‐1 (MALAT1), which was expected to trigger the mTOR pathway through miR‐181b and miR‐144, was differentially expressed in these EVs (Gao et al. 2020). On the other level, aberrant mTOR signaling can modulate EV secretion patterns and cargo selection, thereby generating EVs that are tilted toward carrying pro‐inflammatory signals, which may intensify gliosis, compromise blood–brain barrier integrity, and trigger neuronal apoptosis. Indeed, mTORC1 inhibition stimulates the release of EVs with pro‐tumorigenic functions in cancer (Fan et al. 2020). This reciprocal regulation establishes a vicious cycle: dysregulated mTOR signaling promotes the release of EVs that further exacerbate neuroinflammatory cascades. At the same time, EV‐mediated signals, such as ncRNAs, continue to modulate mTOR activity in a manner that sustains or amplifies the inflammatory response (see Zeng et al. ²⁰²² for a review). The outcome of this bidirectional interaction is a shift in the balance between neuroprotective and neurotoxic processes, with the potential for deleterious effects on neural network function and overall CNS integrity. This layered complexity illustrates how EV‐mediated crosstalk can modulate multiple intracellular pathways, ultimately governing the balance between neuroinflammation and neuroprotection.

4. EVs as Regulators of Neurodegeneration

Every year, neurodegenerative diseases are gaining millions of new patients worldwide. Neurodegeneration is characterized by the progressive death of neurons in both the peripheral and central nervous systems, which leads to motor and cognitive dysfunctions (Wilson et al. 2023). Among the common features of neurodegenerative diseases, neuroinflammation stands out as a central hallmark (Kwon and Koh 2020; Wilson et al. ²⁰²³; Shi and Yong ²⁰²⁵).

As previously stated, EVs can also contribute to the dissemination of toxic molecules responsible for causing neural cell death and neurodegeneration, such as Aβ and α‐syn. Recently, we have identified another mechanism of toxicity transmission via EVs. RTP801 is a stress‐induced protein that acts as a negative regulator of mTOR signaling and has been linked to neurodegeneration and neuroinflammation in different paradigms, such as AD (Pérez‐Sisqués et al. ²⁰²¹), HD (Martín‐Flores et al. ^{2016, 2020}), and PD (Malagelada et al. 2006). RTP801, also known as REDD1, is a stress‐responsive protein that downregulates the mTOR and Akt signaling axis in a neuronal context (Shoshani et al. 2002; Brugarolas et al. ²⁰⁰⁴; Malagelada et al. ²⁰⁰⁸). Even though RTP801 mainly affects mTORC1 activity by maintaining TSC1/2 activity (Brugarolas et al. 2004), evidence from some studies suggests that under specific conditions, RTP801 can influence mTORC2 signaling via disruption of negative feedback loops and modulation of Akt phosphorylation dynamics (Malagelada et al. 2008; Britto et al. ²⁰²⁰). Indeed, overexpression of RTP801 is both necessary and sufficient to trigger neuronal death in vitro. In this context, we have demonstrated that RTP801‐induced toxicity can propagate trans‐neuronally via EVs. Specifically, RTP801 overexpression increases EV release and alters the molecular cargo of neuronal EVs, shifting it toward a pro‐apoptotic profile. In this line, exposure of naïve neurons to these EVs reduces neurite complexity and ultimately triggers apoptosis. On the other hand, EVs derived from cultured neurons where we previously knocked down RTP801 are trophic and prevent recipient neuronal cultures from nutrient deprivation (Solana‐Balaguer, Campoy‐Campos, et al. ²⁰²³; Solana‐Balaguer, Martín‐Flores, et al. ²⁰²³).

The underlying mechanism of neuroinflammation involves the activation of resident immune cells, such as microglia and astrocytes. EVs can cross the CNS barrier, and via EV‐mediated communication, immune signals impact the barrier function and produce a neuroinflammatory response. In this context, EVs serve as key mediators, as they facilitate the communication between immune cells and neurons (Cabrera‐Pastor ²⁰²⁴).

Toll‐like receptors (TLRs) are present in various cells of the nervous system, where they play a role in initiating the inflammatory response. They are activated by signals released from damaged or stressed cells, hence triggering the inflammatory response associated with neuroinflammatory (Izquierdo‐Altarejos et al. ²⁰²³). For instance, the activation of microglial TLRs triggers the release of EVs containing proinflammatory cytokines and chemokines, such as TNF‐α, IL‐1β, caspases, etc. These signals are then propagated to the neighboring cells through EVs, hence amplifying the neuroinflammatory response (Qi et al. 2023).

As previously stated, in neurodegenerative diseases, EVs have been related to both inflammatory conditions and neurodegeneration per se. In the following lines, we aim to explore the intersection between EVs, mTOR signaling, and neuroinflammation in different neurodegenerative conditions.

4.1. Alzheimer's Disease

AD is the most common type of dementia affecting millions of people worldwide. It is characterized by progressive cognitive impairment, along with emotional and psychiatric symptoms. AD is characterized by the presence of neurofibrillary tangles and neuritic plaques, formed by the hyperphosphorylation of Tau protein and the aggregation of Aβ peptides, respectively (DeTure and Dickson ²⁰¹⁹).

EVs are implicated in AD pathogenesis. On one hand, studies have established the trans‐neuronal spread of Aβ and Tau via EVs. Indeed, EVs can induce aggregation of these proteins and propagate their pathology to neurons in vitro (Sardar Sinha et al. 2018) and in vivo (Elsherbini et al. 2020). EVs may facilitate Aβ aggregation and impair its clearance by astrocytes and microglia, contributing to neuroinflammation and cellular dysfunction. Microglia in AD release increased levels of Tau‐containing EVs, promoting Tau spread and disease progression (Dinkins et al. 2014). On the other hand, EVs can also mediate the propagation of neuroinflammatory mediators in AD. For instance, NDEVs and ADEVs isolated from blood samples of AD patients cause the activation of the complement pathway on the surface of primary and iPSC‐derived neurons, cause membrane disruption and neurite fragmentation, and decrease cell viability (Nogueras‐Ortiz et al. ²⁰²⁰). Indeed, plasma‐derived EVs from AD patients show increased expression of the pro‐inflammatory markers IL‐1β, TNFα, and GFAP when compared to the mild cognitive impairment group and control subjects (Singh et al. 2024). A different study showed that excessively Aβ‐accumulating astrocytes in a co‐culture setting generate high phospho‐tau (p‐tau) levels, which trigger the production of ADEVs that ultimately cause neuronal death in AD cells and animal models (Söllvander et al. 2016).

Nonetheless, only a few studies highlight the importance of the mTOR pathway in EVs in AD. EVs are small enough to cross the BBB, suggesting that they may be effective vehicles for the delivery of drugs to the brain. Ebrahim et al. (2024) focus on mesenchymal stem cell‐derived EVs (MSC‐EVs) and their utility for AD treatment. In this study, an aluminum chloride (AlCl₃)‐induced rat model of AD was treated with MSC‐EVs. AlCl₃ treatment in rats induced higher levels of mTOR and p‐Akt. However, the rats treated with MSC‐EVs showed decreased levels of mTOR and p‐Akt, improved cognition, reduced Aβ and p‐Tau accumulation, elevated neurogenesis, enhanced synaptic function, and improved astrogliosis. They related all the aforementioned effects to the modulation of the PI3K/Akt/mTOR signaling pathway possibly via sphingosine 1‐phosphate (S1P) (Ebrahim et al. 2024). Another interesting paper linking Akt signaling and EVs showed that curcumin‐primed EVs potently ameliorate cognitive function in a murine model of AD by inhibiting hyperphosphorylation of the Tau protein through the AKT/GSK‐3β pathway (Wang et al. 2019).

A different study, carried out in the SH‐SY5Y neuroblastoma cell line as a model with different genetic variants of mTOR, highlights that upregulating mTOR activity in vitro promoted Tau localization inside EVs (Tang et al. 2015).

4.2. Parkinson's Disease

PD diagnosis represents a challenge, as motor symptoms appear after the death of most dopaminergic neurons (Cheng et al. 2010). In fact, the prodromal phase of PD can be up to 20 years (Hustad and Aasly 2020).

Many EV studies in PD are mainly focused on finding biomarkers inside EVs isolated from serum, plasma, urine, and saliva. Overall, total α‐syn in neuronal L1CAM‐positive EVs isolated from plasma might be a promising biomarker (Kluge et al. 2022). The same biomarker in saliva EVs is still being investigated, and results converge with previous studies (Kang et al. 2016). Additionally, higher amounts of Ser(P)‐1292‐LRRK2 were found in urinal EVs from idiopathic PD patients (Fraser et al. 2016).

Therefore, EVs and PD are inherently linked. Some forms of PD are caused by mutations at the leucine‐rich repeat kinase 2 (LRRK2) gene and are associated with alterations in the Rab protein family. Rab proteins are crucial modulators of membrane trafficking and orchestrate cell physiology through temporally and spatially regulating docking, fusion, fission, tethering, and vesicle sorting. The role of Rab proteins in PD has been described elsewhere (Bellucci et al. 2022), but still a promising field of study emerges from this intriguing intersection in PD and EVs production.

Indeed, different pathogenic roles of EVs have been described in PD. For instance, Grey and coworkers indicate that exosomes offer an optimal setting for α‐syn to assemble, which may facilitate the spread of PD pathology, despite the modest amounts of α‐syn seen within them (Marie et al. 2015). Indeed, Xia et al. (2021) showed that α‐syn and microglial TLR2 contribute to excessive α‐syn phagocytosis and microglial activation, which then exacerbates the propagation and spreading of α‐syn pathology. Moreover, pro‐inflammatory cytokines increase the accumulation of α‐syn in neurons after microglial EVs containing α‐syn (Guo et al. 2020). Even though the authors do not provide a direct link with the mTOR pathway, microglial EVs containing α‐syn disrupt the autophagic flux via pellino E3 ubiquitin protein ligase 1 (PELI1)‐mediated lysosome‐associated membrane protein 2 (LAMP2) degradation. PELI1 is an E3 ubiquitin ligase abundantly expressed in activated microglia and plays an important role in their inflammatory reaction. Interestingly, PELI1 controls mTOR by inactivating Rheb and suppresses the phosphorylation of mTORC1 target proteins ribosomal protein S6 kinase (S6K) and S6 in CD8+ T cells (Yan et al. 2023).

Other results regarding PD and EVs were reviewed in detail in Li et al. (2022) and Huang et al. (2023). Our group has recently linked mTOR signaling and EVs in an in vitro model of PD. 6‐Hydroxydopamine (6‐OHDA) treatment increases the release of EVs from cortical neurons with elevated levels of the stress‐inducible protein RTP801. These EVs fail to activate the pro‐survival mTOR/Akt signaling pathway in target neurons, and this inhibitory effect is specifically mediated by RTP801. Contrarily, RTP801 knockdown in donor neurons rescues EV‐mediated mTOR/Akt stimulation. These findings indicate a novel mechanism of transcellular toxicity whereby RTP801‐positive EVs propagate stress signals and may contribute to neurodegeneration in PD (Solana‐Balaguer, Campoy‐Campos, et al. ²⁰²³; Solana‐Balaguer, Martín‐Flores, et al. ²⁰²³).

5. Conclusions

In summary, mTOR functions as a central integrator of critical cellular processes that govern neuronal function, survival, and the inflammatory responses in the CNS. The bidirectional crosstalk between mTOR and extracellular vesicles adds an additional layer of regulatory complexity, where EVs act not only as conveyors of regulatory molecules that modulate mTOR activity in recipient cells but also as targets of mTOR‐mediated regulation that determines their release and cargo composition (Makrygianni and Chrousos 2023). This intricate interplay underscores the potential for innovative therapeutic strategies that combine mTOR modulation with targeted EV engineering to recalibrate neuroinflammatory responses and mitigate neurotoxicity. The integration of mTOR inhibitors with agents that modify EV biogenesis or cargo content represents a particularly promising strategy, as it has the potential to jointly attenuate deleterious intracellular signaling and correct aberrant intercellular communication (Zou et al. 2019).

Future studies should aim to delineate the exact molecular mechanisms by which mTOR signaling modulates EV cargo sorting and by which EV‐delivered signals in turn influence mTOR activity in distinct neural cell populations.

In conclusion, the intricate interplay between mTOR signaling, its crosstalk with inflammatory pathways, and the modulatory role of extracellular vesicles constitutes a central nexus in the regulation of neuroinflammation and neurotoxicity. Targeting this multifaceted network through both direct pharmacological modulation of mTOR and strategic manipulation of EV cargo offers a promising avenue for achieving robust neuroprotective effects and improved clinical outcomes in CNS disorders (Hodges and Lugo 2020; Bockaert and Marin ²⁰¹⁵). Continued research into this bidirectional regulatory system will not only expand our fundamental understanding of CNS signaling networks but also pave the way for innovative, integrative therapeutic strategies that address both intracellular and intercellular drivers of neuroinflammation and neurodegeneration.

Author Contributions

P. Garcia‐Segura: conceptualization, writing – review and editing, writing – original draft, investigation, supervision. A. Chicote‐González: conceptualization, writing – original draft, investigation, writing – review and editing. A. Espasa‐Marco: writing – original draft, writing – review and editing. M. Garcia‐Alcaraz: writing – original draft, writing – review and editing. L. Mora‐Bernabé: writing – original draft, writing – review and editing. M. Abelló‐Fernández: writing – original draft, writing – review and editing. C. Malagelada: conceptualization, investigation, funding acquisition, writing – original draft, writing – review and editing, supervision.

Conflicts of Interest

The authors declare no conflicts of interest.

Peer Review

The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer‐review/10.1111/jnc.70256.

Garcia‐Segura, P. , Chicote‐González A., Espasa‐Marco A., et al. 2025. “Tiny Messengers, Huge Consequences: Extracellular Vesicles and mTOR Signaling in Neuroinflammation.” Journal of Neurochemistry 169, no. 10: e70256. 10.1111/jnc.70256.

Data Availability Statement

The authors have nothing to report.