Brain-derived extracellular vesicles: A promising avenue for Parkinson’s disease pathogenesis, diagnosis, and treatment

Abstract

The misfolding, aggregation, and deposition of alpha-synuclein into Lewy bodies are pivotal events that trigger pathological changes in Parkinson’s disease. Extracellular vesicles are nanosized lipid-bilayer vesicles secreted by cells that play a crucial role in intercellular communication due to their diverse cargo. Among these, brain-derived extracellular vesicles, which are secreted by various brain cells such as neurons, glial cells, and Schwann cells, have garnered increasing attention. They serve as a promising tool for elucidating Parkinson’s disease pathogenesis and for advancing diagnostic and therapeutic strategies. This review highlights the recent advancements in our understanding of brain-derived extracellular vesicles released into the blood and their role in the pathogenesis of Parkinson’s disease, with specific emphasis on their involvement in the aggregation and spread of alpha-synuclein. Brain-derived extracellular vesicles contribute to disease progression through multiple mechanisms, including autophagy-lysosome dysfunction, neuroinflammation, and oxidative stress, collectively driving neurodegeneration in Parkinson’s disease. Their application in Parkinson’s disease diagnosis is a primary focus of this review. Recent studies have demonstrated that brain-derived extracellular vesicles can be isolated from peripheral blood samples, as they carry α-synuclein and other key biomarkers such as DJ-1 and various microRNAs. These findings highlight the potential of brain-derived extracellular vesicles, not only for the early diagnosis of Parkinson’s disease but also for disease progression monitoring and differential diagnosis. Additionally, an overview of explorations into the potential therapeutic applications of brain-derived extracellular vesicles for Parkinson’s disease is provided. Therapeutic strategies targeting brain-derived extracellular vesicles involve modulating the release and uptake of pathological alpha-synuclein -containing brain-derived extracellular vesicles to inhibit the spread of the protein. Moreover, brain-derived extracellular vesicles show immense promise as therapeutic delivery vehicles capable of transporting drugs into the central nervous system. Importantly, brain-derived extracellular vesicles also play a crucial role in neural regeneration by promoting neuronal protection, supporting axonal regeneration, and facilitating myelin repair, further enhancing their therapeutic potential in Parkinson’s disease and other neurological disorders. Further clarification is needed of the methods for identifying and extracting brain-derived extracellular vesicles, and large-scale cohort studies are necessary to validate the accuracy and specificity of these biomarkers. Future research should focus on systematically elucidating the unique mechanistic roles of brain-derived extracellular vesicles, as well as their distinct advantages in the clinical translation of methods for early detection and therapeutic development.

Introduction

Parkinson’s disease (PD) is the second most prevalent neurodegenerative disorder (Jovanovic et al., 2025; Wang et al., 2025), with estimated prevalences ranging from 2% to 3% in populations of over 65 years old (Poewe et al., 2017). According to the Global Burden of Disease Study, the incidence of PD is projected to double from 6 million in 2015 to 12 million by 2040, suggesting a potential “PD pandemic” is on the horizon (Dorsey and Bloem, 2018; GBD 2017 Causes of Death Collaborators, 2018). PD is clinically characterized by resting tremors, bradykinesia, myotonia, and postural balance disorders, though patients can also exhibit non-motor symptoms, including autonomic dysfunction, sleep disorders, sensory disturbances, cognitive impairments, and psychiatric disturbances. Key pathophysiological features of PD include the progressive loss of dopaminergic neurons in the substantia nigra (SN) and the accumulation of alpha-synuclein (α-syn) in residual neurons within the affected regions (Kalia and Lang, 2015; Lin et al., 2025; Qi et al., 2025).

In current clinical practice, PD presents two primary challenges. The first is the limited availability of therapeutic options, and the second is the difficulty of early diagnosis. When PD patients present with significant motor dysfunction, they have already experienced a substantial degeneration of dopaminergic neurons in the SN, which can affect up to 50% of neurons, and a marked drop in dopamine transmitter levels in the striatum, often exceeding a 70%–80% decrease (Kalia and Lang, 2015). Due to our limited understanding of the pathogenesis of PD, currently used treatments need improvement. Most available treatments alleviate symptoms without stopping the progression of the disease or reversing extant brain lesions, and they often cause adverse outcomes such as dyskinesia (Bloem et al., 2021). The etiology and mechanisms of PD remain incompletely understood, leading to a lack of accurate biomarkers for early diagnosis, methods for monitoring disease progression, and effective disease-modifying therapies.

With a prodromal period of up to 20 years, PD presents a substantial window for potential neuroprotection and disease intervention if early diagnosis can be achieved (Kalia and Lang, 2015). However, the current lack of accurate early diagnostic methods for PD leaves clinical diagnosis primarily reliant on observable symptoms and imaging techniques (Tolosa et al., 2021). The dopamine transporter protein positron emission tomography (PET) brain phenomenon technique is extensively utilized to identify a reduced dopaminergic nerve terminal density in the basal ganglia. Despite this, it is equally likely to show abnormal results in patients with other dopamine dysfunction disorders unrelated to neuronal α-syn pathology, which limits its diagnostic specificity (Zou et al., 2016; Parnetti et al., 2019). Radiologically-based Machine Learning has been shown to have high specificity and sensitivity, enabling it to differentiate between PD and atypical PD syndrome, along with other similar conditions. Nonetheless, the practical implementation of Machine Learning in PD diagnosis lacks standardized operational guidelines (Makarious et al., 2022). While clinical symptom assessment scales are semi-quantitative and can reflect disease progression to some extent, their reliability is influenced by various uncertainties, including the patient’s condition and the examiner’s experience. Identifying PD in a clinical setting depends on characteristic symptoms. However, typical symptoms often manifest in the later stages, while early non-motor symptoms have low specificity (Tolosa et al., 2009). Some data suggest that 15% of individuals identified with PD fail to satisfy stringent diagnostic standards, and 20% of patients are not diagnosed promptly (Beach and Adler, 2018). Definitive diagnoses can only be established through post-mortem biopsy, emphasizing the urgent need for more effective diagnostic methods. Hence, precise and reliable biomarkers are crucial for differential diagnosis, the monitoring of disease progression, and evaluating treatment efficacy.

The design of α-syn seed amplification assays (SAAs) is based on the observation that pathological α-syn triggers normal α-syn misfolding and aggregation, enabling the specific diagnosis of PD via the analysis of body fluids. A recent study highlighted the potential of using patient skin samples in SAAs. However, obtaining skin samples can be technically challenging, and α-syn seeds present in the skin are highly fragile and susceptible to degradation in non-ultra-cold environments. (Kuang et al., 2024). Moreover, α-syn SAAs have limited effectiveness in distinguishing between clinical PD phenotypes (Siderowf et al., 2023). Cerebrospinal fluid (CSF) levels of α-syn can directly reflect the neurological pathology, but the invasiveness of methods for obtaining CSF makes this approach impractical in most clinical settings. Hence, plasma, blood, and saliva should be explored as alternatives. Brain-derived extracellular vesicles (BDEVs), capable of being transferred from the CSF to the peripheral blood, show promise as potential indicators of PD (Upadhya and Shetty, 2021; Herman et al., 2023).

Extracellular vesicles (EVs), nanoscale particles secreted by cells, contain proteins, nucleic acids, and lipids and play key roles in both physiological and pathological processes. Given their ability to transport biomarkers and cross the blood–brain barrier (BBB), EVs have recently garnered considerable attention for their diagnostic and therapeutic potential in neurodegenerative diseases (Banks et al., 2020). BDEVs are a specialized class of EVs originating from cells within the nervous system, such as neurons, glial cells, Schwann cells, and other neural-associated cells (Figure 1). A significant subset of BDEVs are the neuron-derived EVs (NDEVs) specifically produced by neurons. NDEVs are typically isolated using methods that rely on neuronal surface markers, such as L1 cell adhesion molecule (L1CAM), that differentiate them from the EVs produced by other brain cell types.

Figure 1.

Relationships between EVs and BDEVs.

EVs can be secreted by nearly all types of cells in the human body. Among these, EVs originating from brain-specific cells are referred to as BDEVs. These include EVs secreted by neurons, microglia, astrocytes, Schwann cells, and other brain cell types. BDEVs have garnered significant attention in PD and other neurodegenerative diseases. Created with BioRender.com. BDEVs: Brain-derived extracellular vesicles; EVs: extracellular vesicles; PD: Parkinson’s disease.

BDEVs can be isolated from peripheral blood, providing a window into CNS lesions. These vesicles play critical roles in multiple pathogenic mechanisms of PD, including neuroinflammation, oxidative stress, and α-syn aggregation and spread. Moreover, BDEVs are closely associated with neural regeneration, underscoring their significance in disease pathogenesis and potential therapeutic applications. Particular attention has been given to NDEVs due to their potential diagnostic prominence in PD. Current clinical studies are primarily focusing on isolating NDEVs from the peripheral blood or CSF to identify disease-specific biomarkers. Key molecules such as α-syn and various miRNAs carried by NDEVs have been validated as biomarkers for the early detection and monitoring of PD. Furthermore, a recent study demonstrated the seeding activity of α-syn within NDEVs using a standardized SAA, highlighting their potential as valuable biomarkers for PRKN-linked PD (Kluge et al., 2024). This finding indicates not only the diagnostic value of NDEVs themselves but also their potential to improve existing diagnostic methods for PD. BDEVs hold significant promise in the therapeutic landscape of PD. These vesicles have been explored as carriers for delivering therapeutic molecules, e.g., drugs, miRNAs, and proteins, across the BBB to target affected neural tissues. Moreover, BDEVs are closely associated with neural regeneration, underscoring their multifaceted applicability in PD treatment.

This review focuses on the role of BDEVs in the pathogenesis, diagnosis, and treatment of PD, with particular attention given to the diagnostic applications of NDEVs. By exploring the unique characteristics and potential of BDEVs, we aimed to provide a comprehensive overview of their significance in advancing our understanding of PD and the development of innovative diagnostic and therapeutic approaches.

Search Strategy

To gather relevant literature, a systematic search was carried out in the PubMed, Scopus, and Web of Science databases for articles published until August 31, 2024. The search strategy incorporated various keywords, including “extracellular vesicles,” “brain-derived extracellular vesicles,” “exosome,” “α-synuclein,” “Parkinson’s disease,” “pathogenesis,” “diagnosis,” “biomarker,” “therapeutics,” and “neural regeneration.” The initial results were screened based on their titles and abstracts to identify studies specifically addressing the role of BDEVs in PD with a focus on their mechanisms, diagnostic potential, therapeutic applications, and contributions to neuroregeneration. Publications that lacked full-text access were excluded from this review. No limitations were imposed on the language of publication or type of study, ensuring that all relevant sources were included in the analysis.

Overview of Extracellular Vesicles

Evolution of extracellular vesicle research

The study of EVs has advanced significantly since the mid-21^st century, evolving from initial morphological observations to detailed functional analyses. In the 1940s, small platelet-derived particles with anticoagulant properties were first identified in plasma, marking the discovery of the first non-cellular components in body fluids (Chargaff and West, 1946). In the 1960s, Wolf (1967) utilized electron microscopy to further characterize these particles, which he referred to as “platelet dust,” and demonstrated their separation via high-speed centrifugation (Wolf, 1967). During the 1970s, Crawford (1971) conducted more detailed studies, revealing that these vesicles contain lipid structures and carry molecules such as ATP. These particles were then termed “microparticles.” By the 1980s, Harding et al. (1983) confirmed that cells release vesicles containing intracellular substances, and they named these vesicles exosomes. Initially, exosomes were believed to function solely as containers for waste expulsion by cells (Johnstone et al., 1987). By the end of the 1990s, it was discovered that EVs are involved in biological signal transduction, and more attention was being paid to their biological functions (Zitvogel et al., 1998).

Since these early studies, further breakthroughs have been made in the role of EVs in neurodegenerative diseases (Bard et al., 2004; Valadi et al., 2007). In 2005, α-syn was first found to be present in EVs, providing new insights into the pathological mechanisms of PD (Lee et al., 2005). In 2014, L1CAM was proposed as a marker for NDEVs and has since been widely used in a study exploring the diagnostic potential of NDEVs in PD (Shi et al., 2014). By 2016, research had demonstrated that exosomes in PD patients’ CSF contained pathogenic α-syn and induced its abnormal aggregation in target cells, emphasizing the critical role of EVs in facilitating pathological transmission between neurons (Stuendl et al., 2016). Cao et al. in 2017 revealed the presence of a range of miRNAs with abnormal expression in the EVs of PD patients, indicating the significant potential of these miRNAs as early diagnostic biomarkers. After decades of exploration, BDEVs are now recognized as important mediators of intercellular communication in the nervous system (Figure 2).

Figure 2.

Timeline of key milestones in EV research.

CSF: Cerebrospinal fluid; EVs: extracellular vesicles; L1CAM: L1 cell adhesion molecule; miRNA: microRNAs; NDEVs: neuron-derived extracellular vesicles; PD: Parkinson’s disease; α-syn: α-synuclein.

Definition, biological roles, and significance of extracellular vesicles

EVs can be classified into three primary subtypes: apoptotic bodies, microvesicles, and exosomes. Apoptotic bodies, which originate from apoptotic cells, have diameters ranging from 1000 nm to 5000 nm. Microvesicles are released from the plasma membrane of cells through clathrin mechanisms and typically range in size from 50 nm to 1000 nm. Exosomes, generally between 30 nm and 150 nm across, are generated via the endocytic pathway or originate from apoptotic bodies (Van Niel et al., 2018; Doyle and Wang, 2019). The classification and nomenclature of EVs follow certain established guidelines but are not fully standardized, with some studies using the general term “EVs” and others focusing specifically on “exosomes.” This review retains the original terminology used in cited studies to ensure accuracy, while using “EVs” as a more general term for broader discussions.

EVs are produced by almost every cell type within the body, such as neurons, microglia, astrocytes, and erythrocytes (Pascua-Maestro et al., 2019; Jiang et al., 2020a), and they play a vital role in intercellular communication. Their membranes are composed of a variety of lipids and membrane proteins, such as tetraspanins, fusion and transfer proteins, lysosome-associated membrane glycoproteins, heat shock proteins, cytoskeletal proteins, integrins, transferrin receptors, and MHC molecules. EVs can be found in different body fluids, e.g., CSF, blood, urine, and saliva. The effective extraction of EVs from the blood and the identification of biomarkers may provide effective methods for risk screening and the early diagnosis of PD patients. Multiple factors, including the generative mechanisms, physiological conditions, type of originating cell, and cellular environment, shape the composition of EVs. The EXOCARTA database has evidence that a wide range of proteins, miRNAs, mRNAs, and lipids can be present in EVs (Yuan et al., 2021). EVs are suggested to be crucial participants in various physiological functions, such as neurogenesis, angiogenesis, proinflammatory and anti-inflammatory responses, cellular apoptosis, and wound healing (Gupta and Pulliam, 2014; Pascual et al., 2020). They are essential for cellular differentiation, the maintenance of tissue homeostasis, and organ remodeling in nearly all tissues, including the CNS. Consequently, EVs provide pivotal pathways for intercellular communication.

EVs have drawn considerable attention among neuroscientists due to their ability to traverse the BBB. EVs in the bloodstream can reflect the pathological state of the CNS and serve as valuable biomarkers, and they can act as therapeutic delivery vehicles. The efficiency of EV BBB crossing remains limited, as studies suggest that EVs are more likely to exit the CNS by moving into peripheral circulation rather than entering the brain. However, an inflammatory environment may facilitate the entry of EVs into the brain through the BBB (Chen et al., 2016; Matsumoto et al., 2017). Furthermore, the specific mechanisms and pathways governing EV transport across the BBB remain poorly understood (Saint-Pol et al., 2020; Ramos-Zaldívar et al., 2022). Therefore, while the potential of EVs to traverse the BBB presents an exciting and rapidly advancing area of research, considerable uncertainties and challenges persist.

Isolation and detection of extracellular vesicles

The isolation and detection of EVs are prerequisites for and the foundations of EV-related research. The detection of blood-derived EVs faces numerous technical challenges, necessitating the further refinement of current methodologies. Multi-omics analyses, including proteomic, genomic, and lipidomic approaches, are critical for identifying potential biomarkers within EVs (Clos‐Garcia et al., 2018).

The existing isolation techniques for EVs warrant further optimization. Currently, standard methods include ultracentrifugation (UC)—the most widely used technique—as well as polymer precipitation, immunoaffinity capture chromatography, microfluidic approaches, and size-exclusion chromatography (Takov et al., 2019). The physical characteristics of EVs, including their size and morphology, are typically analyzed using imaging technologies such as transmission electron microscopy, scanning electron microscopy, and atomic force microscopy. Although commercial kits for EV extraction are now available, methods for the accurate quantification of EVs are significantly limited, as current measurement methods, such as protein quantification and nanoparticle tracking analysis, remain imprecise. In addition, commonly used methods for identifying the contents of EVs, especially RNA and proteins, include quantitative RT-PCR, high-throughput sequencing, western blotting, and ELISA. Despite the availability of these assays, the successful development of EVs-based PD biomarkers requires standardised and reproducible protocols to improve the accuracy and efficiency of EVs isolation and characterisation.

Ultracentrifugation remains the most widely used and cost-effective method for EV isolation. However, it often produces samples with impurities that thus require additional purification steps (Muraoka et al., 2022). NDEVs are typically isolated using immunocapture techniques targeting L1CAM, a transmembrane protein regarded as a neuronal marker. Nevertheless, the specificity of L1CAM has come under increasing scrutiny. L1CAM is also expressed in non-neuronal cells and circulates in body fluids as soluble fragments due to cleavage or alternative splicing (Angiolini et al., 2019). This can lead to the unintended capture of soluble L1CAM fragments, which lack neuron-specific EV content. Furthermore, studies have demonstrated that elevated α-syn levels in exosomes isolated using anti-L1CAM antibodies may result, in part, from non-specific interactions between soluble α-syn and the antibodies in plasma (Norman et al., 2021). These challenges highlight the necessity of identifying more specific biomarkers for isolating brain-derived EVs.

To overcome these limitations, brain-specific proteins identified from EV proteomic datasets have been proposed as potential alternatives to L1CAM. Proteins such as STXBP1, GPM6A, PSD2, and GDI1 have shown strong intercorrelated expression in brain tissues and may serve as promising nervous system-specific markers (Inoue et al., 2015; Sharma et al., 2015). However, further investigations and rigorous clinical validation are required to confirm their specificity and reliability for NDEV isolation. In summary, isolation techniques will need to be refined and brain-specific biomarkers identified to improve the specificity, purity, and reproducibility of NDEVs for clinical applications, particularly in PD. Future advancements in isolation technologies and analytical methodologies will be necessary for enhancing the reliability of molecular cargo analysis.

Roles of brain-derived extracellular vesicles in the nervous system

BDEVs are important for intercellular communication in the nervous system. They mediate the transfer of proteins, lipids, and nucleic acids between neurons, glial cells, and other brain-resident cells. BDEVs play crucial roles in the nervous system throughout an individual’s life, impacting processes ranging from CNS development and the maintenance of cerebral homeostatic resistance to pathological changes in the aging brain (Huo et al., 2021).

BDEVs are intricately linked to synaptic plasticity and have been found to contain specific miRNAs, such as Let7c and miR21, that regulate dendritic growth (Liu et al., 2015). They are associated with neuroprotective mechanisms such as neural regeneration, anti-stress responses, and cellular waste removal. After spinal cord injury, BDEVs transport miR-124-3p, which suppresses A1 astrocytes and aids in injury repair through modulation of the miR-124-3p/MYH9 pathway (Jiang et al., 2020b). Moreover, the EVs induced by heat shock in neural stem cells exhibit substantial neuroprotective properties by counteracting the oxidative stress and neurotoxicity caused by amyloid-beta proteins (Huber et al., 2022). Another important function of BDEVs is their clearance of intracellular materials. When the autophagic–lysosomal pathway is impaired, EVs can assist in eliminating toxic protein aggregates. The effective release of EVs contributes to removing endolysosomal material from cells, regulating flux through neuronal endosomal pathways, and reducing transport stress. The stabilized or increased production of BDEVs may serve as a protective response, potentially preventing or mitigating the neuropathological endolysosomal alterations associated with aging and neurodegenerative diseases (Mathews and Levy, 2019).

Oligodendrocytes, responsible for synthesizing and maintaining CNS myelin sheaths, provide electrical insulation to axons and nutritional support to neurons. EVs derived from oligodendrocytes are crucial in facilitating these functions, as they contain proteins essential for myelin formation, neuronal nutrition support, synaptic plasticity, and antigen presentation (Sherman and Brophy, 2005; Nave et al., 2023). Astrocytes are involved in vital CNS functions, such as forming and maintaining the BBB, modulating neural plasticity, and regulating neurotransmission and metabolism, that are often mediated through EVs (Sofroniew, 2020). Thus, EVs are essential for CNS development, safeguarding neurons, aiding recovery, and modulating neuronal functions, but they also play significant roles in the initiation and progression of CNS diseases.

Role of Brain-Derived Extracellular Vesicles in the Pathogenesis of Parkinson’s Disease

Brain-derived extracellular vesicles in the transmission of pathogenic proteins

EVs act as carriers of neurotoxic substances; an example of which is α-syn, which exhibits prion-like behavior, enabling its spread from affected to healthy neurons (Dehay et al., 2015; Goedert, 2015). The dissemination of α-syn is enabled through various mechanisms, including endocytosis, nanotube tunneling, exosomes, and glymphatic flow (Goedert, 2015). BDEVs play a pivotal role in the long-distance dispersion of α-syn (Alvarez-Erviti et al., 2011a). Previous studies have demonstrated that α-syn-overexpressing SH-SY5Y cells release EVs containing α-syn, which are then taken up by normal SH-SY5Y cells, contributing to their pathological involvement. Aberrant α-syn secretion via EVs, rather than its direct cellular release, may enhance its uptake by recipient cells (Gustafsson et al., 2018), as the α-syn associated with EVs is readily internalized by neurons. Another study showed that recipient cells internalized α-syn oligomers associated with EVs 2.4 times more effectively than their free-floating counterparts, with a marked decrease in uptake observed when the integrity of the exosomal membrane was compromised. When bound to EVs, these oligomers triggered greater toxicity, as demonstrated by significant caspase 3/7 activation and increased apoptosis in both neuronal and cellular models. These findings highlight the role of EVs as specialized vectors that mediate intercellular α-syn transfer and amplify the pathological effects of α-syn in PD (Danzer et al., 2012). In PD models, neurons harboring aggregated α-syn are a key source of EVs containing phosphorylated α-syn. These NDEVs can be internalized by recipient neurons via endocytosis. Following uptake, the vesicles merge with the endosomal system of the host cell and are transported along axonal pathways to interconnected downstream neurons. During this process, host endosomes partially degrade the vesicular contents while also re-secreting them, enhancing both the dissemination and pathological effects of α-syn (Polanco et al., 2018). Research has shown that α-syn can be exported from neurogenic cells through EVs and released into the surrounding matrix (Emmanouilidou et al., 2010). Another study has revealed that extracellular α-syn can induce excessive microglial activation by stimulating the purinergic P2X7 receptor, leading to neurotoxicity. These two cellular events may be sequential steps in the pathological cascade of neurodegeneration.

BDEVs significantly influence the aggregation, conformation, and spread of α-syn. Whereas NDEVs contribute to autophagy-lysosomal dysfunction, a critical factor in the progression of PD, as impairments in this pathway exacerbate the accumulation of toxic protein aggregates. A recent study demonstrated that the EVs derived from α-syn-overexpressing SH-SY5Y cells exhibit significantly elevated levels of miR-19a-3p. These EVs can be taken up by recipient microglia, leading to an increase in miR-19a-3p expression within these cells. MiR-19a-3p has been shown to inhibit autophagy by targeting the PTEN/AKT/mTOR signaling pathway, resulting in impaired lysosomal function and reduced degradation of pathological α-syn (Zhou et al., 2019). Investigations using BV2 microglial cells have shown that microglia readily internalize plasma exosomes from patients with PD, which compromises their autophagic machinery. This process leads to both an enhanced accumulation of intracellular α-syn and increased secretion of α-syn into the extracellular space (Xia et al., 2019). Additionally, microglia exposed to human α-syn preformed fibrils exhibit impaired autophagic processes, resulting in the release of EVs enriched with α-syn. These vesicles effectively promote protein aggregation in recipient neurons (Guo et al., 2020). Additionally, research indicates that NDEVs carry pathological α-syn, characterized by β-sheet and fibrillary conformations, which are detectable in blood samples and contribute to the pathological aggregation of α-syn within neuronal networks. The conformation-specific antibody MJFR-14-6-4-2 can precisely recognize these β-sheet and fibrillary forms of α-syn. Researchers using in vitro seeding assays further demonstrated that NDEVs containing pathological α-syn significantly enhanced the aggregation rate of recombinant monomeric α-syn, resulting in the formation of amyloid fibrils. Furthermore, using repeated cycles of seeding amplification, the researchers revealed that NDEVs from PD patients exhibited substantially increased seeding activity, with the amplified products displaying fibrillary conformations, as confirmed by transmission electron microscopy. Interestingly, although total α-syn levels did not differ significantly between patients and healthy individuals, the pathological conformations of α-syn were present at notably higher levels in PD patients (Kluge et al., 2022). Additionally, research indicated that EVs extracted from the CSF of individuals with α-syn-related neurodegeneration can promote the oligomerization of soluble α-syn in vitro (Jiang et al., 2020a). Astrocyte-derived EVs (ADEVs) are also closely involved in the mechanisms of α-syn propagation and neuronal injury, and dopaminergic neurons efficiently internalize ADEVs. However, the EVs derived from LRRK2 G2019S mutant astrocytes were shown to exhibit diminished neurotrophic support. This reduction was linked to the accumulation of phosphorylated α-syn and other pathological proteins within the multivesicular bodies (MVBs) and EVs of mutant astrocytes. After being taken up by neurons, these ADEVs failed to maintain neuronal health and instead promoted dendritic atrophy and neuronal dysfunction. Additionally, EVs from LRRK2 G2019S astrocytes enhanced the spread of LRRK2 and phosphorylated α-syn between cells, aggravating stress responses and degeneration in dopaminergic neurons via non-cell-autonomous pathways (De Rus Jacquet et al., 2021).

Conversely, α-syn has been demonstrated to have a reciprocal relationship with the generation and release of EVs, establishing EVs as essential carriers for α-syn propagation (Emmanouilidou et al., 2010). Autophagic dysfunction results in impaired α-syn secretion, causing the elevated release of α-syn-rich EVs. Upon internalization, the degradation of α-syn through the endosomal sorting complex required for transport (ESCRT) pathway within MVBs is impaired due to reduced levels of charged multivesicular body protein 2B (CHMP2B), leading to the impaired internalization and abnormal accumulation of α-syn. Notably, reduced CHMP2B levels are a characteristic feature of synucleinopathies. Gene therapy targeting lentivector-CHMP2B has demonstrated efficacy in reducing α-syn accumulation and improving neurodegenerative pathology in transgenic mice (Spencer et al., 2016). Research also suggests that inhibiting autophagy by downregulating Baf or SNAP29 enhances EV release. α-Syn was shown to suppress the expression of SNAP29, thereby increasing the release of EVs. These EVs may facilitate the propagation of pathological proteins toward neighboring neurons. Nanoparticle tracking analysis confirmed that α-syn and rapamycin stimulated the release of EVs in the cell culture media (Tang et al., 2021). One study using primary astrocytes with A53T α-syn overexpression or exposure to aggregated α-syn showed a significant increase in ADEV secretion. This process was associated with impaired lysosomal activity, as evidenced by the reduction in cathepsin L activity and lower levels of lysosome-associated membrane proteins. The accumulation of undigested autophagosomes suggested that α-syn aggregation disrupted the autophagy-lysosomal pathway and thus contributed to increased ADEV release (Wang et al., 2023). These studies suggest that pathological α-syn promotes the production and release of brain-derived EVs, facilitating intercellular communication and contributing to the progression of PD.

Brain-derived extracellular vesicles in neuroinflammation

Inflammatory processes are crucial to the pathogenesis of PD. Increasing evidence indicates that neuroinflammation originating in the periphery enhances central inflammatory responses. This process activates microglia and leads to astrocyte atrophy, which not only worsens neurodegenerative conditions but also amplifies the detrimental effects of α-syn oligomers (Xu et al., 2018). The strong correlation between peripheral EVs and neuroinflammation, which intensifies with age, is evidenced by changes in the concentration of EVs in senescent cells. For instance, transferring EVs from 24-month-old mice to 3-month-old mice significantly activated neuroglial cells in the young mice (Morales-Prieto et al., 2022). Research has indicated that EVs originating from inflammatory macrophages can induce extensive neuroinflammation by activating microglia and astrocytes, as well as upregulating proinflammatory cytokines. Comprehensive analyses involving sequencing, bioinformatics, and functional assays have identified miR-155-5p as a pivotal element for triggering inflammatory responses within glial cells (Li et al., 2016; Jin et al., 2023b).

Glial cell-derived EVs are essential for mediating interactions between neuroglial cells and neurons, contributing to neuroinflammation (Ibáñez et al., 2019). For instance, glial cell-derived EVs carry chemokines that can bind to Toll-like receptor 2 (TLR2) and Toll-like receptor 4 (TLR4) on neurons, triggering neuroinflammation (Gupta and Pulliam, 2014). Additionally, ADEVs are known to transport pathogenic proteins with misfolded conformations, dysregulated microRNAs, and proinflammatory cytokines, which can be transferred to healthy neurons, promoting the spread of neuroinflammation (Gupta and Pulliam, 2014). The NLRP3 inflammasome, predominantly expressed in microglia, is a critical component of neuroinflammation. EVs carrying α-syn can activate NLRP3 expression, initiating the neuroinflammatory response. Moreover, EVs formed during NLRP3 assembly can encapsulate proinflammatory cytokines such as IL-1β and IL-18 and thereby amplify neuroinflammation (Cypryk et al., 2017, 2018). The presence of Mn²⁺ is also recognized to be a contributing factor to PD. Prolonged exposure to Mn²⁺ markedly elevates NLRP3 inflammasome expression in microglia. Furthermore, the EVs released by microglia primed with LPS and exposed to Mn²⁺ are enriched with apoptosis-associated speck-like protein containing a CARD (ASC), a critical inflammasome component, and are more prolific under these conditions. These EVs are subsequently taken up by nearby microglia, further enhancing NLRP3 inflammasome activation and IL-1β secretion (Sarkar et al., 2019). The EVs derived from manganese-stimulated dopaminergic neurons carrying pathological α-syn aggregates are internalized by microglia via caveolin-1-mediated endocytosis. This process induces microglial activation, characterized by the release of proinflammatory cytokines such as necrosis factor alpha (TNF)-α and interleukin (IL)-1β and IL-6 (Harischandra et al., 2019). These altered EVs possess potent capabilities for intercellular transport, which can significantly impact neuroinflammation, exacerbating the condition. Microglia-derived EVs (MDEVs) have been implicated in promoting neuroinflammation and dopaminergic neurodegeneration. A study revealed that microglia from aged mice showed a diminished ability to phagocytose exosome-bound oligomeric α-syn compared to those from younger mice, while concurrently exhibiting an amplified inflammatory response marked by increased TNF-α production (Bliederhaeuser et al., 2016). In midbrain slice cultures, inflammatory stimulation with IFN-γ and LPS not only activated microglia but also significantly increased the rate of MDEV release. The addition of these MDEVs to untreated cultures directly induced a loss of dopamine neurons, suggesting their direct neurodegenerative and inflammatory effects were independent of microglial activation (Tsutsumi et al., 2019). Plasma-derived EVs isolated from PD patients, which are enriched with oligomeric α-syn, are preferentially internalized by microglia both in vitro and in vivo. Following uptake, these EVs activate microglia, inducing a reactive state characterized by the release of proinflammatory cytokines, such as TNF-α and IL-6, along with elevated nitric oxide production. This microglial activation not only triggers an inflammatory response but also enhances the secretion of α-syn via EVs (Xia et al., 2021). In summary, the inflammatory response is an important mechanism for the BDEV-mediated promotion of PD.

Brain-derived extracellular vesicles in oxidative stress

NDEVs are involved in the development of PD through oxidative stress-related mechanisms. DJ-1, known for its antioxidative properties in neurons, is regulated post-transcriptionally by hsa-miR-4639-5p; this microRNA targets the 3′-untranslated region of DJ-1, resulting in reduced DJ-1 expression. Thus, elevated hsa-miR-4639-5p expression leads to decreased DJ-1 levels, heightening oxidative stress and causing neuronal damage. Neuronal functional impairment in neurodegenerative diseases often involves disrupted synaptic transmission linked to increased reactive oxygen species synthesis and decreased synaptic transmission efficiency (Zhang et al., 2014; Kamat et al., 2016). Elevated hsa-miR-4639-5p levels in a patient’s plasma can serve as a diagnostic indicator. Studies suggest that this plasma miRNA may originate from NDEVs released by the CNS, which explains its elevated expression in PD patients’ plasma (Chen et al., 2017). NDEVs can be isolated using immunocapture with the biomarkers L1CAM and NCAM1, and hsa-miR-4639-5p is markedly overexpressed in these exosomes compared to exosome-free plasma (Shi et al., 2014; He et al., 2023). In PD, reduced circSV2b expression in EVs leads to decreased Foxk1 expression, which subsequently reduces Akt1 levels and intensifies oxidative stress-induced damage. Drugs for upregulating circSV2b expression have potential benefits for PD diagnostics and therapeutics, as they have been reported to aid in restoring dopamine synthesis and improving motor function in mouse models (Cheng et al., 2022; Figure 3).

Figure 3.

Role of BDEVs in PD.

BDEVs play a critical role in the progression of PD through mechanisms such as α-syn aggregation and propagation, neuroinflammation, and oxidative stress. Pathological α-syn can be transmitted from damaged neurons to healthy ones via BDEVs, facilitating the abnormal spread of this protein. EVs carrying misfolded α-syn, dysregulated microRNAs, and inflammatory mediators can infiltrate glial cells, triggering their activation and driving neuroinflammation. In turn, activated astrocytes and microglia release EVs enriched with inflammatory mediators and neurotoxic substances, further amplifying the inflammatory cascade and contributing to neurotoxic injury. Additionally, BDEVs promote the pathogenesis and progression of PD by inducing oxidative stress. Created with BioRender.com. BDEVs: Brain-derived extracellular vesicles; EVs: extracellular vesicles; GCase: glucocerebrosidase; miRNA: microRNAs; PD: Parkinson’s disease; ROS: reactive oxygen species; α-syn: α-synuclein.

Brain-Derived Extracellular Vesicles as Biomarkers for Diagnosis of Parkinson’s Disease

Comparison between brain-derived extracellular vesicle analysis and other diagnostic methods for Parkinson’s disease

The diagnostic landscape for PD currently includes standardized rating scales, neuroimaging techniques, fluid- and tissue-based α-syn biomarkers, and genetic testing. These approaches have advanced PD diagnosis, but each has distinct advantages and limitations. Clinical assessments are practical and cost-effective for evaluating motor symptoms but lack specificity, especially in the early stages. Neuroimaging techniques such as PET and MRI are employed to detect brain changes but are costly and poorly accessible. Molecular biomarkers such as α-syn require invasive sampling of the CSF, serum, or tissue, limiting patient compliance. While genetic testing can detect PD-related mutations, it is limited to familial cases, emphasizing the need for minimally invasive and reliable diagnostic tools. The analysis of BDEVs shows great promise as an alternative strategy for diagnosing PD.

Standardized rating scales remain essential for PD diagnosis and monitoring. The Unified PD Rating Scale is widely used to assess motor and non-motor symptoms, disease severity, and functional impairment. Other commonly used tools include the Hoehn and Yahr staging system, which classifies disease progression into five stages based on motor dysfunction, and the Non-Motor Symptoms Scale, specifically designed to evaluate non-motor symptoms such as sleep disturbances, cognitive impairment, and mood disorders. These scales are applied to effectively assess PD progression but are influenced by inter-rater variability and the rater’s clinical experience.

Imaging techniques such as dopamine transporter single-photon emission computed tomography (DAT-SPECT) and MRI are widely used to detect structural and functional changes in the brain (Tatsch and Poepperl, 2013; Armstrong and Okun, 2020). Advanced MRI techniques, including neuromelanin imaging, quantitative susceptibility mapping, and dorsal nigral hyperintensity evaluation, offer promising tools for studying nigral pathology (Wang et al., 2019). However, their high cost and limited accessibility restrict their applicability in early diagnosis and routine monitoring.

Fluid and tissue α-syn markers have garnered significant attention among PD researchers (Andréasson and Svenningsson, 2021). The α-syn levels in CSF are considered reliable indicators of PD. However, the invasiveness of lumbar puncture limits sampling repetitions. Blood-based α-syn assays, while minimally invasive, face challenges due to presence of α-syn in peripheral red blood cells, which can confound measurements (Parnetti et al., 2019). Genetic testing has identified SNCA, LRRK2, and GBA gene mutations as significant risk factors for familial and sporadic PD. However, these mutations do not consistently predict the onset, progression, or severity of sporadic PD, which accounts for most cases (Ye et al., 2023).

Compared to these methods, BDEVs offer unique advantages. BDEVs can be isolated from peripheral blood in a minimally invasive and repeatable manner, providing access to neuron-specific biomarkers, including α-syn, miRNAs, and other molecular cargo. These vesicles overcome the limitations of whole-blood α-syn assays by specifically reflecting CNS pathology. BDEVs offer the potential for the early diagnosis and longitudinal monitoring of disease progression and therapeutic responses and provide a specific and accessible complementary approach.

NDEVs represent a significant and specialized subset of BDEVs and have garnered particular attention in studies on the diagnosis of PD. NDEVs provide a direct link to neural activity, while general EVs derived from a heterogeneous mix of cells lack this specificity. In healthy individuals, NDEVs demonstrate significantly elevated levels of neuron-specific proteins when compared to other EVs subtypes, such as CD81⁺ EVs. Quantitative assessments indicate that essential neuronal markers, including L1CAM, NSE, NFL, NCAM, and proBDNF, are enriched by a factor of two to four in NDEVs relative to those in CD81⁺ EVs. Additionally, NDEVs contain markedly higher concentrations of neuronal proteins, such as phosphorylated tau and BDNF, than plasma, serum, or total EV populations. Interestingly, these enriched neuronal markers, which are often challenging to detect in non-enriched EV populations, are reliably identified in NDEVs, highlighting their sensitivity and specificity for reflecting neuronal pathophysiological changes. The associations between NDEV-linked α-syn and phosphorylated tau and disease severity further support their potential as biomarkers for neurodegenerative disorders, including Alzheimer’s disease (AD) and PD. Furthermore, NDEVs display distinct protein profiles indicative of metabolic activity, differentiating them from other EV subtypes. Proteins involved in cellular energy regulation and signaling, such as mTOR, Akt, and the leptin receptor, are present at higher levels in NDEVs, reflecting the heightened metabolic activity of their neuronal origin. These distinctive molecular characteristics suggest that BDEVs are particularly advantageous for studying brain-specific pathophysiology (Mustapic et al., 2017).

Although comparative studies on the differences between NDEVs and general EVs in PD have been limited, investigations in AD have highlighted the significant distinctions among their origin, molecular composition, and functional roles. For example, it has been shown that the miRNA profiles of NDEVs exhibit a unique signature distinct from total EVs. While in one study, 59% of miRNAs were found to be unique to total EVs and 26% were exclusive to NDEVs, only 15% overlapped between the two populations. NDEVs are enriched with CNS-specific miRNAs, such as miR-23a-3p, miR-223-3p, and miR-190a-5p, which are upregulated in AD, while miR-100-3p is downregulated. These findings emphasize the unique potential of NDEVs as reflections of CNS-specific pathological changes. Functional analyses further demonstrated that NDEVs are enriched in neural pathways, including axon guidance and synaptic regulation, highlighting their specificity for CNS conditions (Serpente et al., 2020). In the context of PD, the pathological aggregation and propagation of α-syn are considered central mechanisms, with inflammation and oxidative stress also playing significant roles. Both NDEVs and other EVs have been identified as carriers of α-syn that facilitate its intercellular transmission. However, it remains unclear whether the α-syn carried by NDEVs exhibits stronger seeding ability than that in other EVs, or whether differences exist in the conformation and content of α-syn within these vesicle subtypes. Furthermore, from a diagnostic perspective, it has been suggested that NDEVs show specific changes in the early stages of PD, potentially providing a method for early diagnosis (Ohmichi et al., 2019). In summary, BDEVs exhibit unique advantages in the diagnosis of neurological disorders, including PD (Table 1).

Table 1.

Comparison of diagnostic techniques for Parkinson's disease

Diagnostic methods	Principles and detection indicators	Advantages	Limitations	Current applications	References
BDEVs	Extract BDEVs from bodily fluids and analyze the specific biomarkers they contain, such as proteins, RNA, and lipids	High sensitivity for detecting early pathological changes; non-invasive or minimally invasive sampling methods; detect multiple molecular biomarkers; Facilitate real-time dynamic monitoring	Lack of standardized extraction and detection methods; complexity of bioinformatics interpretation	Research stage with ongoing exploration of diagnostic potential in PD	Natale et al., 2022
MRI	Evaluates structural brain changes through high-resolution imaging to exclude other diseases with similar symptoms	Anatomical imaging; enables direct observation of structural or functional brain changes; eliminates the need for radioactive tracers	Limited in reflecting molecular pathological mechanisms; low sensitivity of conventional MRI for early-stage PD	Routine diagnostic tool for detecting brain structural changes or excluding other diseases	Wang et al., 2019
DAT-SPECT	Assesses the integrity of nigrostriatal dopaminergic neurons using specific radiotracers targeting dopamine transporters	Functional imaging; detection of functional deficits in dopaminergic neurons; facilitates early diagnosis	Low disease specificity; risk of radiation exposure; high cost	Routine diagnostic tool for functional diagnosis of PD and other movement disorders, particularly in suspected early cases	Tatsch and Poepperl, 2013
Fluid and tissue-based α-synuclein	Measures α-synuclein levels in bodily fluids or tissues	Reflects changes in the core pathogenic protein α-synuclein; simple sampling technique	Invasive sampling procedure; sensitivity and specificity require further validation	Research phase with no routine clinical application	Andréasson and Svenningsson, 2021
Clinical rating scales	Comprehensive assessment of patient symptoms using clinical rating scales	Non-invasive and easy-to-perform procedure; comprehensive assessment of symptom severity; enables monitoring of disease progression	Results influenced by subjectivity, dependent on evaluator’s expertise; inability to reflect early molecular or structural changes of the disease	Core tool for clinical diagnosis, typically requires combination with other methods for differential diagnosis	Bloem et al., 2021
Genetic testing	Detects Parkinson’s disease-related genetic mutations such as LRRK2, PINK1, SNCA, and GBA	Clarifies genetic factors; high specificity for diagnosing familial PD; applicable for risk screening	Limited to a subset of hereditary cases; low diagnostic value for sporadic PD	High diagnostic value in familial PD cases	Ye et al., 2023

BDEVs: Brain-derived extracellular vesicles; DAT-SPECT: dopamine transporter single-photon emission computed tomography; MRI: magnetic resonance imaging; PD: Parkinson’s disease.

Protein biomarkers for Parkinson’s disease diagnosis

α-Synuclein in brain-derived extracellular vesicles: diagnostic value and detection strategies

α-Syn is an essential biomarker in PD and exerts a substantial effect on its pathogenesis. Lewy bodies (LBs), which are characteristic of PD, comprise aggregates of α-syn. Under normal physiological homeostasis, α-syn is produced during neurogenic phenotype determination and synaptic connection establishment. However, its expression diminishes in instances of neuronal damage or synaptic plasticity impairment. Pathologically, misfolded and aggregated α-syn exhibits neurotoxic properties. It is predominantly located at neuronal synapses and interacts intricately, influencing synaptic activity by modulating neurotransmitter release and synaptic vesicle aggregation and recycling (Sharma and Burré, 2023).

The detection of α-syn has been a prominent focus in neuropathological research on PD. A definitive criterion for PD diagnosis is the detection of LBs formed from misfolded α-syn within nigrostriatal dopaminergic neurons in the brain, although this method is not feasible in clinical practice. Tissues, such as the intestine, salivary glands and skin, and body fluids, including CSF, urine, and blood, are potential sources of α-syn. Researchers found no significant difference in the signal intensity of soluble α-syn in NDEVs in PD individuals compared to healthy people. However, pathological α-syn was substantially heightened in PD and could be detected by MJFR-14-6-4-2, a structure-specific antibody targeting the β-sheet-rich structures of α-syn, including oligomers, protofibrils, and insoluble fibrils, which are the main pathogenic forms of the protein. A study found that “[¹⁸F]-F0502B” could be used to visualize α-syn aggregations in neuronal cells in synucleinopathy cases and thus serve as a promising new PET tracer. The parallel diagonal stacking of F0502B can occur on the surface of α-syn fibers through ligand-ligand interactions (Xiang et al., 2023). Structure-specific antibodies against α-syn are effective in differentiating PD from healthy tissue (Kluge et al., 2022), and different strains of α-syn can be amplified and detected using protein misfolding cyclic amplification (Shahnawaz et al., 2020). Real-time quaking-induced conversion is a diagnostic tool based on the identification of α-syn aggregations that rapidly amplifies template-mediated misfolded protein aggregates in CSF. However, blood contains inhibitory substances that can obstruct the seeding amplification of aggregated proteins, posing challenges for blood-based real-time quaking-induced conversion (Martinez-Valbuena et al., 2022). One strategy used in developing sensitive and effective biomarkers focuses on fluid components other than soluble proteins, such as EVs. The Quanterix Simoa technique, which surpasses the accuracy of traditional ELISA, involves quantifying L1CAM-exosomal α-syn in plasma at sub-pg/mL concentrations. Commercial assay kits are also effective for detecting α-syn. A meta-analysis showed that the α-syn within L1CAM EVs extracted from blood samples using the ExoQuick kit and measured via ELISA exhibits potential as an early indicator for PD (Nila et al., 2022). ELISA and western blotting are standard methods for α-syn detection. Using western blotting, the pronounced elevation in α-syn levels within NDEVs isolated from the blood of all PD patients can be demonstrated. Moreover, α-syn seeding assays have validated the presence of β-sheet-rich structures and a fibril-like organization, confirming the pathological folding of α-syn (Kluge et al., 2022). Another study using ELISA, Luminex assays, mass spectrometry, and multiplex electrochemiluminescence detected elevated levels of α-syn in NDEVs from PD patients in comparison to those from healthy controls (Jiang et al., 2020a). EVs play a crucial role in the early diagnosis, determination of disease course and severity, and differential diagnosis of PD. Blood, being easily obtainable and rich in EVs, serves as an ideal sample for such analyses. Notably, neural-derived EVs provide greater precision in their reflection of the condition of the CNS.

α-Synuclein in brain-derived extracellular vesicles used to distinguish patients with Parkinson’s disease from healthy controls

The α-syn in BDEVs serves as a valid marker for differentiating PD from healthy individuals. PD patients have lower concentrations of α-syn in their CSF but significantly higher levels of plasma L1CAM EVs than healthy people. The total levels of DJ-1 and α-syn do not differ significantly between healthy and PD individuals; nonetheless, the concentrations of DJ-1 and α-syn in NDEVs are markedly elevated in individuals with PD, and there is a positive correlation between these biomarkers. Another elevated indicator is the ratio of DJ-1 in plasma NDEVs to total DJ-1 (Zhao et al., 2019), markers that are applied to determine diagnosis rather than disease progression. A 2021 study highlighted the reliability of Ser129 phosphorylated α-syn (p-α-syn) and oligomeric α-syn for PD diagnosis. PD patients exhibited substantially increased ratios of oligomeric to total α-syn and p-α-syn within plasma EVs compared to healthy people, making these markers useful for PD diagnosis (Zheng et al., 2021).

Most studies have indicated that α-syn levels become elevated in blood neurogenic EVs as PD progresses. However, some studies have suggested there is a negative correlation between progression and α-syn levels in plasma EVs in PD patients with akinetic-rigid subtype motor symptoms. Reduced plasma EV α-syn levels in PD may be due to the toxic α-syn aggregates entrapped in neurons, which hinder α-syn monomer loading into EVs and inhibit their release (Duce et al., 2017). The chelation of α-syn into Lewy vesicle protofibrillar aggregates and the increased neuronal uptake of α-syn could decrease α-syn levels in both CSF and plasma EVs (Lööv et al., 2016; Hijaz and Volpicelli-Daley, 2020). The possible erythrocytic origin of EVs carrying α-syn nucleoprotein in the blood may lead to biased data interpretation. When investigating blood EVs, it is essential to clearly define the sample type, the EVs source, and the specific type and conformation of α-syn of interest.

α-Synuclein in brain-derived extracellular vesicles for adjunctive diagnosis of Parkinson’s disease subtypes

Patients with PD may exhibit different motor symptoms, and examining exosomal α-syn can help differentiate between different movement subtypes. A previous study detected α-syn concentrations in NDEVs in the serum of PD and essential tremor patients. The average measurement observed in PD was lower than that in both the essential tremor patients and healthy control groups. Among the PD cohort, levels were higher in the tremor-dominant group than the non-tremor-dominant group (Si et al., 2019). Rapid eye movement sleep behavior disorder (RBD) is a frequent non-motor manifestation in individuals with PD. Patients with PD and RBD exhibit higher levels of oligomeric α-syn in NDEVs, elevated total scores on the RBD screening questionnaire, and more severe clinical symptoms. Furthermore, oligomeric α-syn levels in NDEVs and soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) complex component concentrations in serum hold promise as reliable biomarkers for RBD-specific PD phenotypes (Meloni et al., 2023a).

Using combinations of α-syn with other biomarkers is essential for improving diagnostic accuracy. Several studies have been undertaken to assess the potential benefits of integrating α-syn with additional biomarkers and to discover more precise and sensitive detection indices. One such inquiry observed the levels of α-syn oligomers and presynaptic SNARE proteins (STX-1A, VAMP-2, and SNAP-25) in EVs derived from the peripheral blood. PD patients exhibited a conspicuous increase in oligomeric α-syn in NDEVs compared to healthy individuals (Agliardi et al., 2021; Meloni et al., 2023a), and this increase was positively correlated with the progression and severity of PD pathology. In contrast, significant reductions in STX-1A and VAMP-2 were noted (Table 2).

Table 2.

NDEVs as biomarkers to distinguish PD from HC

Biomarkers	Sources	Isolation of EVs	Biomarker detection	Data analysis and statistics	Levels in PD	Conclusions	Participants	References
α-Syn	Plasma L1CAM exosomes	Exosomes were isolated from mouse or human plasma using antibody-coated superparamagnetic microbeads	An established Luminex protocol	AUC = 0.654, 70.1% sensitivity, 52.9% specificity	Higher	Compared with HC, patients with PD have lower concentrations of α-syn in cerebrospinal fluid but significantly higher levels of α-syn in plasma L1CAM exosomes.	267 PD, 215 HC	Shi et al., 2014
a-Syn	Plasma neural cell–derived exosomes	Exosomes were isolated according to the experimental protocol described in a previous report	ELISA kits	AUC = 0.654, 48.7% sensitivity, 85.0% specificity	Higher	The levels of DJ-1 and α-syn in plasma neural-derived exosomes were significantly elevated in PD patients, and there was a positive correlation between the two biomarkers.	39 PD, 39 HC	Zhao et al., 2018
DJ-1	Plasma neural cell–derived exosomes	Exosomes were isolated according to the experimental protocol described in a previous report	ELISA kits	AUC = 0.703, 79.5% sensitivity, 57.5% specificity	Higher	The ratio of plasma neural- derived exosomes of DJ-1 to total DJ-1 was significantly elevated; The concentrations of DJ-1 in the plasma neural-derived exosomes were significantly higher in patients with PD compared with HC.	39 PD, 39 HC	Zhao et al., 2018
Ser129 phosphorylated α-syn	Plasma L1CAM exosomes	Exosome precipitation method	Western blot	α-Syn oligomer/total α-syn in exosomes AUC = 0.71, 60.5% sensitivity, 59.4% specificity; the ratio of p-α-syn oligomer/total p-α-syn AUC = 0.69, 60.0% sensitivity, 59.5% specificity	higher	The ratios of oligomeric α-syn/total α-syn and oligomeric p-α-syn/total p-α-syn in the plasma exosomes of patients with PD were significantly higher.	36 PD, 36 HC	Zheng et al., 2021
α-Syn	Plasma neural cell–derived exosomes	Antibody-coated superparamagnetic microbeads following a protocol adapted from Tauro was used to isolate NDEVs	ELISA kit	AUC = 0.675, 66.7% sensitivity, 71.1% specificity	Lower	The levels of α-syn in CNS-derived exosomes in the serum of PD is lower than in the ET and HC groups.	38 PD (22 TD and 16 NTD), 21 ET, and 18 HC	Si et al., 2019
α-Syn oligomers and presynaptic SNARE proteins	Neural cell–derived EVs in blood	Immunocapture with biotinylated L1CAM antibody	ELISA kit	NDEVs oligo α-syn AUC = 0.824, 78.1% sensitivity, 75.0% specificity; NDEVs STX-1A AUC = 0.753, 75.0% sensitivity, 77.5% specificity; VAMP2 AUC = 0.780, 81.3% sensitivity, 72.5% specificity	Higher/ lower	Oligomeric α-syn was significantly augmented whereas STX-1A and VAMP-2 were significantly reduced in NDEVs of PD patients compared to HC.	32 PD, 40 HC,	Agliardi et al., 2021
miRNA	Exosome-like microvesicles in serum	Total exosome isolation reagent from body fluids	qRT‐PCR	sensitivity and specificity values: miR-195, 82.6% and 55.0%; miR-19b, 68.8% and 77.5%; and miR-24, 81.7% and 85.0%	Higher/lower	This study validated the downregulation of miR-19b and the upregulation of miR-195 and miR-24 in patients with PD.	109 PD, 40 HC	Cao et al., 2017
miRNA	Exosomes in serum	ExoQuick: The exosome pellet was directly lysed with Qiazol Lysis Reagent for total RNA. isolation.	miRNA expression was analyzed using real-time PCR.	Univariate ROC curves were computed for the 12 significant miRNAs analyzed in the expanded cohort.	Higher/lower	MiRNA signatures could non-invasively discriminate between neurodegenerative disorders; specifically, let-7d, miR-15b, miR-24, miR-142-3p, miR-181c, and miR-222 appear to be associated with Parkinson-like phenotypes.	30 AD, 30 PD, 24 VD, and 25 VP	Barbagallo et al., 2020
Pro-IL-1β TNF-α	Plasma EVs	exoEasy Maxi kit	Western blot analysis of cytokines in plasma EVs	The nonparametric Mann–Whitney U test and Spearman’s rank correlation	Higher	Patients with PD had significantly elevated levels of plasma EV pro-IL-1β and TNF-α compared to HC. Additionally, plasma EV pro-IL-1β, IL-6, IL-10, and TNF-α were associated with the severity of cognitive dysfunction.	113 PD, 48 HC	Chan et al., 2021

AUC: Area under the curve; CNS: central nervous system; EVs: extracellular vesicles; ELISA: enzyme-linked immunosorbent assay; HC: healthy control; IL-1β: interleukin-1β; L1CAM: L1 cell adhesion molecule; NDEVs: neuron-derived extracellular vesicles; PCR: polymerase chain reaction; PD: Parkinson’s disease; qRT‐PCR: quantitative reverse transcription-polymerase chain reaction; ROC: Receiver Operating Characteristic; TNF-α: tumor necrosis factor-alpha; VAMP2: vesicle-associated membrane protein 2; α-syn: α-synuclein.

α-Synuclein as a biomarker for differential diagnosis of Parkinson’s disease

In the early stages of PD, the symptoms often overlap with those of other neurological disorders. Atypical Parkinsonism is a syndrome that shares similarities with PD, such as motor impairment, tremors, and muscle tone abnormalities, but also presents atypical symptoms. Atypical Parkinsonism encompasses progressive supranuclear palsy (PSP), multiple system atrophy (MSA), corticobasal ganglionic degeneration, and dementia with LBs (Levin et al., 2016). Furthermore, distinguishing PD from AD, primary lateral sclerosis, and other neurological diseases is essential. Biomarkers are necessary for distinguishing between these conditions due to the overlap in symptoms.

α-Syn plays a crucial pathological role in both PD and MSA, but it predominantly aggregates in neurons in PD and oligodendrocytes in MSA. The comparative concentration of α-syn between oligodendrocytes and neurons acts as a precise marker for distinguishing PD from MSA (Reddy and Dieriks, 2022). In blood NDEVs, α-syn levels are significantly higher in PD than controls but lower than those observed in MSA (Dutta et al., 2021). Research has indicated there is an increase in α-syn concentrations in NDEVs isolated from serum during the prodromal stage of PD, with levels progressively increasing throughout the disease course, suggesting the progressive impairment of dopaminergic neurons is caused by α-syn. A single cross-sectional measurement in four cohorts (PD, MSA, PSP, and CBS) demonstrated that α-syn concentrations in L1CAM EVs exceeding 14 pg/ml could be used to differentiate between PD and MSA (Jiang et al., 2021). Exosomal α-syn has been shown to have superior diagnostic efficacy compared to total CSF α-syn (Eusebi et al., 2017). A recent investigation on serum-derived EVs from individuals with PD, MSA, PSP, RBD, and a control group revealed markedly elevated filamentous α-syn within EVs in PD compared to HC, MSA, and PSP samples (Ishiguro et al., 2024). Analysis of NDEVs revealed significantly increased oligomeric α-syn and decreased Tau aggregates in PD compared to atypical Parkinsonism. The levels of Tau aggregates and oligomeric α-syn in NDEVs have emerged as the best biomarkers for diagnosing PD/PSP and PD/corticobasal ganglionic degeneration (Meloni et al., 2023b). Using astrocyte-specific markers such as GLT-1, researchers isolated ADEVs and found that those in the plasma of PD patients contained significantly higher levels of total and aggregated α-syn than those of healthy individuals and MSA patients. ROC analysis further validated the diagnostic potential of ADEVs, demonstrating their high sensitivity and specificity in differentiating PD from MSA and healthy individuals and underscoring their value in differential diagnosis (Wang et al., 2023; Table 3).

Table 3.

NDEVs as biomarkers for differential diagnosis of PD

Biomarkers	Sources	Isolation of EVs	Biomarker detection	Data analysis and statistics	Levels in PD	Conclusions	Participants	References
Filamentous α-syn	Serum NDEVs	Immunoprecipitation using L1CAM	ELISA	Tukey’s multiple comparisons test or Sidak's multiple comparisons test	Higher	A significant reduction in the number of EVs was observed in patients with PD compared to controls. Furthermore, there was a substantial increase in filamentous α-syn within EVs in patients with PD when compared to controls, as well as those with MSA and PSP.	142 PD, 18 MSA, 28 PSP	Ishiguro et al., 2024
Oligomeric α-syn	Serum NDEVs	A two-step method was adapted from a previously published procedure.	ELISA	PD and APS AUC = 0.817, 78.6% sensitivity, 77.5% specificity, cut off value > 0.637 ng/mL	Higher	Oligomeric α-synuclein was significantly elevated in PD compared to APS.	70 PD, 21 PSP, 19 CBD	Meloni et al., 2023b
αSyn	Serum L1CAM exosomes	Poly(carboxybetaine-methacrylate)-coated magnetic beads were used to isolate L1CAM-positive exosomes.	Triplexed electrochemiluminescence was used to measure exosomal α-synuclein.	PD and MSA AUC = 0.98, 87% sensitivity, and 97% specificity; PD and HC AUC = 0.86, 77% sensitivity, and 74% specificity	Higher	The α-syn content in L1CAM-positive exosomes was elevated by approximately 2-fold in PD compared to HC, MSA, or other tauopathies.	PD 290, MSA 50, PSP 116, CBS 88, HC191	Jiang et al., 2021
α-Syn	NDEVs	Exosomes were isolated from serum or plasma using immunoprecipitation with neuronal and oligodendroglial markers.	Electrochemiluminescence ELISA	Separated PD from MSA with an AUC = 0.902, 89.8% sensitivity and 86.0% specificity	Higher	α-Syn concentrations in blood exosomes of neurogenic origin were significantly higher in patients with PD compared to HC, but lower than those found in MSA.	50 HC, 51 PD, 30 MSA	Dutta et al., 2021
miRNAs	Serum	miRNAs were extracted using the Qiagen miRNeasy Mini Kit, following the Qiagen supplementary protocol for the purification of small RNAs from serum and plasma.	TaqMan Low Density Array technology	Expression profiles were analyzed using TaqMan Low Density Arrays, along with single TaqMan assays, and the Wilcoxon rank-sum test was applied for statistical analysis.	Higher/lower	The expression of miR-24, miR-34b, and miR-148b was upregulated in the serum of patients with MSA, whereas miR-148b and miR-30c were downregulated in patients with PD.	25 PD, 25 MSA, 25 HC	Vallelunga et al., 2014

AUC: Area under the curve; CBD: corticobasal degeneration; CBS: corticobasal syndrome; ELISA: enzyme-linked immunosorbent assay; EVs: extracellular vesicles; L1CAM: L1 cell adhesion molecule; MSA: multiple system atrophy; NDEVs: neuron-derived extracellular vesicles; PD: Parkinson’s disease; PSP: progressive supranuclear palsy; α-syn: α-synuclein.

Potential of α-synuclein for early diagnosis and implications for disease monitoring

The diagnostic potential of α-syn within NDEVs in the blood is substantial for various stages of PD. In one study, patients with early PD were prospectively followed to document α-syn levels and clinical symptom changes. Elevated levels of α-syn over time, rather than initial baseline levels, were linked to a heightened risk of motor progression observed over an average span of 22 months. The α-syn levels in NDEVs correlated with the 16-item Sniffin’ Sticks test (SS-16), Unified Parkinson’s Disease Rating Scale part III (UPDRS-III), and Nonmotor Symptom Questionnaire (NMSQ) scores of patients with PD (Niu et al., 2020). In serum samples, α-syn oligomers within NDEVs showed positive correlations with indicators of PD progression, such as disease duration, UPDRS motor scores, and modified Hoehn and Yahr Scale. In contrast, NDEV-Tau aggregate concentrations were negatively correlated with overall cognitive scores (Meloni et al., 2023b).

RBD can be a prodromal symptom of PD. In a study that compared plasma total exosome α-syn (t-exo) and neurogenic exosome α-syn (n-exo) levels in PD, possible isolated RBD (pRBD), and healthy groups, the researchers found significant elevations in both t-exo and n-exo levels in PD patients, whereas individuals with pRBD exhibited an increase solely in n-exo α-syn expression (Yan et al., 2022). A cross-sectional study involving 576 participants identified L1CAM-positive EVs containing α-synuclein in serum as a promising diagnostic biomarker for PD. The study established a critical threshold of 17.75 pg/mL, which distinguishes individuals with over an 80% probability of having prodromal PD from those with less than a 5% risk (Yan et al., 2024). Additionally, a negative correlation was also found between the α-syn content in L1CAM-positive EVs and CSF by α-syn seed amplification assay. In conclusion, evaluating the levels of α-syn in EVs may provide a valuable method for identifying PD pathology in individuals at elevated risk of the condition (Yan et al., 2024).

In a clinical study, researchers collected plasma from 267 PD patients and 215 controls and confirmed that α-syn in plasma NDEVs had comparable sensitivity and specificity to α-syn in CSF for determining PD severity (Shi et al., 2014). Research employing Quanterix Simoa to quantify α-syn revealed a substantial positive relationship between Linc-POU3F3 and α-syn in the NDEVs of PD patients, who had elevated levels compared to controls. The correlation between the β-glucocerebrosidase and α-syn proteostasis has also been documented. Long non-coding RNAs (lncRNAs) play pivotal roles in the autophagy-lysosome pathway by inhibiting autophagy and are stabilized in plasma EVs (Guo et al., 2015). β-Glucocerebrosidase activity was decreased and negatively correlated with Linc-POU3F3. The levels of these biomarkers in PD vary significantly with gender, modified Hoehn and Yahr Scale stage, and UPDRS-III distribution and can be used in combination for PD diagnosis and severity assessment (Zou et al., 2020). PD plasma exosome α-syn is linked to the severity of ankylosis symptoms, and α-syn in tissue or body fluids can directly assist with PD diagnosis (Chung et al., 2021b). Serum α-syn changes markedly with disease progression and helps with the determination of PD progression but cannot be used to differentiate between PD and healthy individuals. However, the presence of α-syn in the skin and submandibular gland is highly specific for PD, offering new possibilities for its early diagnosis (Chahine et al., 2020; Table 4).

Table 4.

NDEVs for early diagnosis and implications for disease monitoring

Biomarkers	Sources	Isolation of EVs	Biomarker detections	Data analysis and statistics	Levels in PD	Conclusions	Participants	References
α-Syn	Plasma EVs	exoEasy Maxi kit	Immunomagnetic reduction assay	AUC = 0.631, 50% sensitivity, 76% specificity	Lower	Plasma EV α-synuclein levels were significantly lower in patients with PD and showed a negative correlation with the severity of motor symptoms in the AR subtype.	116 PD, 46 HC	Chung et al., 2021b
α-Syn	Plasma neuron-derived exosomes	Previously reported protocols	Electrochemiluminescence immunoassays	The optimal cutoff of α-syn for distinguishing patients with early-stage PD from HCs was 395.34 pg/mL, with an AUC of 0.8, demonstrating 100% sensitivity and 57.1% specificity. Spearman correlation analysis in the total PD patient cohort revealed that α-syn was significantly correlated with both UPDRS III motor scores and the total UPDRS (I + II + III) scores.	Higher	The α-syn levels in plasma neuronal exosomes were significantly higher in patients with early-stage PD compared to HCs. Additionally, longitudinal increases in α-syn levels, rather than baseline levels, were associated with a higher risk of motor symptom progression in PD. The study included 36 patients with early-stage PD, 17 patients with advanced PD, 20 individuals with iRBD and 21 HCs.	36 early-stage PD, 17 advanced PD, 20 iRBD, 21 HC	Niu et al., 2020
α-Syn	Plasma L1CAM exosomes	Antibody-coated superparamagnetic microbeads	Quanterix Simoa	AUC = 0.616, with 55% sensitivity, 61% specificity	Higher	Linc-POU3F3 and α-syn concentrations in L1CAM exosomes were increased, while GCase activity was decreased in patients with PD compared to HCs. Significant correlations were observed among L1CAM exosomal Linc-POU3F3 levels, GCase activity, and the severity of PD, including both motor and cognitive dysfunction.	93 PD, 85 HC	Zou et al., 2020
α-Syn	Plasma neural-derived exosomes	L1CAM-coated Dynabeads	Western blotting	t-exo α-syn AUC = 0.741, 75.0% sensitivity and 68.8% specificity n-exo α-syn AUC = 0.761, 77.3% sensitivity and 66.7% specificity.	Higher	Both t-exo and n-exo were significantly increased in patients with PD, while patients with iRBD showed only increased expression of n-exo α-syn.	78 PD, 153 probable iRBD and 63 HC	Yan et al., 2022
Neurofilament light chain	Plasma-derived EV	exoEasy Maxi Kit	Immunomagnetic reduction assay	Setting the cutoff value of plasma NfL at 12.34 pg/mL, the sensitivity and specificity of the PD prediction model were 53.2% and 90.5%, respectively.	Higher	There was no difference in NfL levels in plasma-derived EVs between patients with PD and HCs; however, there was a trend toward increased NfL levels in PD patients with severe motor ankylosis.	116 PD, 46 HC	Chung et al., 2020a
BDNF	Plasma EV	exoEasy Maxi Kits	Enzyme-linked immunosorbent assay	A nonparametric Mann-Whitney U test was performed to compare the plasma exosomal levels of BDNF and other continuous variables between patients with PD and HC.	Lower	Plasma EV BDNF was significantly associated with motor function.	114 PD, 42 HC	Chung et al., 2020b
Cytokines	Plasma EVs	An exoEasy Maxi kit	Western blotting	Spearman’s rank correlation was used to assess the association between plasma EV cytokine levels and the UPDRS scores, as well as cognitive test outcomes.	Higher	The severity of cognitive dysfunction was significantly associated with plasma EV levels of pro-IL-1β, IL-6, IL-10, and TNF-α.	113 PD, with mild to moderate stage disease, 48 HC	Chan et al., 2021
tau and Aβ1–42	Plasma exosomes	exoEasy Maxi kit	Immunomagnetic reduction-based immunoassay	Spearman’s rank-order correlation was performed to investigate the association between plasma EV tau and Aβ1–42 levels and cognitive function.	Higher	High levels of tau and Aβ1–42 in plasma exosomes are significantly associated with cognitive dysfunction in PD.	116 PD, 46 HC	Chung et al., 2021a
miRNA	Plasma brain-derived exosomes	Transferrin-conjugated magnetic nanoparticles	Quantitative polymerase chain reaction	Undescribed	Higher	miR-195-5p, miR-495-3p, miR-323a-3p, and miR-30b-5p displayed increasing upregulation with advanced 2.5-4 stage in PD.	20 PD, 5 HC	Jang et al., 2024

AR: Akinetic-rigid; AUC: area under the curve; Aβ1–42: amyloid-beta 1–42; BDNF: brain-derived neurotrophic factor; EVs: extracellular vesicles; HCs: healthy controls; IL-10: interleukin-10; IL-6: interleukin-6; iRBD: idiopathic rapid eye movement sleep behavior disorder; L1CAM: L1 cell adhesion molecule; NfL: neurofilament light chain; PD: Parkinson’s disease; pro-IL-1β: pro-interleukin-1beta; TNF-α: tumor necrosis factor-alpha; UPDRS: Unified Parkinson’s Disease Rating Scale; α-syn: α-synuclein.

Use of DJ-1 in brain-derived extracellular vesicles as a biomarker

Characterized by its antioxidant properties, DJ-1 is crucial for safeguarding neurons and other cells from neurotoxicity and oxidative stress, primarily through the modulation of the Nrf2 system (Lind-Holm Mogensen et al., 2023). DJ-1 confers its protective benefits by means of a spectrum of pathogenic mechanisms, such as its ability to maintain mitochondrial integrity, influence redox signaling kinase pathways, and function as a regulatory element in transcription (Gallinat et al., 2022). Recent research has provided evidence that DJ-1 mitigates the glycation and aggregation of α-syn, alleviating oxidative stress in PD. Therefore, impaired DJ-1 function, due to mutations or oxidative stress, results in increased glycated α-syn, exacerbating PD pathology (Sharma et al., 2019). The analysis indicated that total DJ-1 and α-syn levels are comparable between controls and PD. However, in PD patients, the concentrations of DJ-1 and α-syn in plasma NDEVs were significantly elevated and positively correlated, and there was an increased ratio of plasma NDEVs DJ-1 to total DJ-1. These markers have been linked to diagnosis but not to disease progression (Zhao et al., 2019).

Evidence suggests that DJ-1 upregulates VMAT2 expression to safeguard dopamine levels (Lev et al., 2013). A short peptide called ND-13, created by fusing a 13-amino acid segment from DJ-1 with a 7-amino acid sequence from TAT, has been found to mitigate the deterioration of DA neurons and improve behavioral deficits in PD mouse models (Lev et al., 2015). In PD patients, oxidized DJ-1 concentrations in urine specimens were reported to be twice as high as those in healthy individuals. While plasma concentrations of DJ-1 were similar between the two groups, the levels of DJ-1 in NDEVs in plasma were notably elevated in individuals with PD. Consequently, DJ-1 has emerged as a viable biomarker for diagnosing PD.

RNA biomarkers for Parkinson’s disease diagnosis

MicroRNAs as biomarkers for Parkinson’s disease

The RNA variants found in EVs encompass mRNAs, tRNAs, circRNAs, and ncRNAs. Among these are also miRNAs, which are highly abundant and diverse in EVs, making them a focus of biomarker research. MiRNAs, a type of non-coding small RNA, consist of 20 to 24 nucleotides. When miRNAs attach to the 3′-UTR of mRNA, they regulate gene expression regulation post-transcriptionally (Bartel, 2009), and several miRNAs have been identified as regulators of α-syn. For instance, the downregulation of miR-34b and miR-34c leads to increased α-syn expression, which is initially associated with mitochondrial dysfunction and oxidative stress and consequently accelerates PD progression (Miñones-Moyano et al., 2011; Kabaria et al., 2015). Elevated levels of miR-16-1 levels suppress HSP70 mRNA translation, leading to reduced α-syn levels. The overexpression of miR-44438 inhibits the production of EVs, causing the hindered efflux and increased aggregation of α-syn in neurons, likely through altered NDST1 mRNA expression. A decrease in heparan sulfate function has also been shown to play a role in this process (Huang et al., 2022). In PD, reduced miR-34b expression in specific brain regions is associated with subsequent decreases in PARKIN and DJ-1 levels.

MiRNAs can enter body fluids in two forms: as ribonucleoprotein complexes with argonaute proteins or as cargo within EVs (Schulz-Siegmund and Aigner, 2021). Exosomal miRNAs are more accurate indicators of cellular state and are more useful for disease diagnosis than free miRNAs (Gurung et al., 2021). Additionally, exosomal miRNAs are resistant to degradation by RNase and remain stable in body fluids, suggesting they have potential as diagnostic biomarkers for PD (Preethi et al., 2022). For example, miR-7-1-5p and miR-223-3p are present at significantly elevated levels in the serum and EVs of PD patients and can be used to distinguish these patients from healthy individuals (Citterio et al., 2023). Moreover, miR-223-3p detrimentally regulates the cytosolic innate immune signaling receptor NOD-, LRR-, and pyrin domain-containing 3 (NLRP3) and modulates the levels of key proteins, thereby affecting inflammasome activation (Mancuso et al., 2019). PD diagnosis can be facilitated by detecting changes in a set of biomarkers in serum EVs in the form of decreased miR-19b and elevated miR-24 and miR-195 levels (Cao et al., 2017). Additionally, lets-7d, miR-15b, miR-142-3p, miR-181c, and miR-222 appear to be associated with a Parkinson-like phenotype (Barbagallo et al., 2020).

MiRNAs in EVs have been demonstrated to be valuable in determining the stage and prognosis of PD. Among various screened miRNA, miR-331-5p have emerged as an indicator for PD disease progression and treatment response (Cardo et al., 2013). Conversely, miR-29c, miR-29a, and miR-19b were shown to be downregulated in PD, but have no significant correlation with AD, hence their simultaneous downregulation is not considered a universal marker of neurodegenerative diseases (Botta-Orfila et al., 2014). The miRNAs k-TSP1 (miR-1826/miR-450b-3p), miR-626, and miR-505 were also identified as sensitive biomarkers for predicting PD, demonstrating high sensitivity, specificity, positive predictive value, and negative predictive value (Khoo et al., 2012). A clinical study employing weighted gene co-expression network analysis and the human miRNA Disease Database and identified 17 miRNAs with specific expression in different stages of PD. After validation, hsa-miR-374a-5p and hsa-miR-374b-5p were reported to be generally upregulated in stages II, III, and IV of PD. Hsa-miR-28-5p was explicitly upregulated in stage III, hsa-miR-22-5p specifically upregulated in stage IV, and hsa-miR-151a-5p was downregulated in stage IV (He et al., 2021). This suggested that miRNAs can not only aid in diagnosing PD but also play a role in clinical staging. However, a limitation of using miRNAs as diagnostic biomarkers is the insufficient depth and precision of studies into specific miRNAs. Prospective cohort studies exploring the association between miRNA levels in circulating EVs and alterations in clinical symptoms will be essential.

MiRNAs contribute to the differential diagnosis of PD. Using TaqMan Low-Density Array technology, a study examined 754 miRNAs and discovered nine circulating miRNAs with differential expression in PD and MSA patients relative to healthy controls. MiR-24, miR-34b, and miR-148b were upregulated in MSA serum, while miR-148b was downregulated in PD (Vallelunga et al., 2014). A study of a Japanese cohort showed that miR-16 differed most significantly among 50 upregulated miRNAs and may be involved in α-syn aggregation (Kume et al., 2018). Genome-wide array analysis of MSA and PD patients revealed miR-7641 and miR-191 to be capable of differentiating between PD and MSA. Additionally, notch signaling, identified as the most relevant pathway after gene network analysis of aberrantly expressed miRNAs, was revealed to play a role in prion disease (Pérez-Soriano et al., 2020)

MiR-23a, miR-29a, and miR-181c are applied in distinguishing PD from vascular parkinsonism, while let-7d, miR-24, miR-142-3p, miR-22*, miR-23a, and miR-222 are used in differentiating PD from HC. However, a single miRNA may be differentially expressed in multiple diseases, necessitating the use of biomarker combinations to enhance the sensitivity and specificity of diagnoses. Moreover, miRNA levels may vary between exosomal and intracellular environments, emphasizing the need for repeated studies of exosomal miRNA levels (Barbagallo et al., 2020).

The diagnostic value of miR-44438 in alpha-synucleinopathies, including PD, has been demonstrated in multicenter clinical studies. Increased levels of miR-44438 were detected in plasma EVs from patients with α-synucleinopathies or RBD compared to healthy individuals, but no similar findings appeared for other conditions. Notably, the diagnostic utility of this miRNA for PD was further reinforced when researchers found a decline in EV levels containing miR-44438 as the disease progressed. However, miR-44438 cannot be used to differentiate between PD and MSA. A later study further optimized the approach and innovated the nano-cage MB assay for the in situ analysis of miRNAs. This nanoscale flow cytometry assay was used for the quantitative detection of EV miRNA levels and enhanced the plausibility of the conclusions (Yu et al., 2024). The utility of miRNAs for diagnosis is being increasingly confirmed in studies, further justifying researchers’ focus on these as biomarkers for PD.

The isolation of total EVs from body fluids is easier to achieve than that of NDEVs. However, the specificity of the differentially expressed miRNAs detected is lower for EVs. For instance, a study involving 31 healthy participants and 72 PD patients at different Hoehn and Yahr stages identified 185 differentially expressed miRNAs in serum EVs. However, only six miRNAs proved to be potential biomarkers (He et al., 2021). It is worth noting that the dysregulation of miRNAs in non-specific EVs mainly reflects the overall body pathophysiology, such as inflammation, immune regulation, and oxidative stress, rather than any physiology specific to PD. For example, miR-34a, implicated in apoptosis and oxidative stress regulation, is overexpressed in the EVs of PD patients, possibly indicating an early inflammatory response in these patients. Elevated levels of miR-34a-5p were detected in small EVs isolated from plasma free of exogenous protein contaminants (Grossi et al., 2021). However, miR-34a-5p is also elevated in other diseases, such as polycystic ovary syndrome and rheumatoid arthritis (Wu et al., 2021; Cui et al., 2024). Similarly, miR-21, which was shown to be differentially expressed in circulating plasma EVs in a pre-motor PD model, is also upregulated in various non-tumor diseases and cancers (Jenike and Halushka, 2021). These findings highlight the challenges in using general EVs as specific diagnostic tools for PD, as multiple factors influence their miRNA expression patterns.

NDEVs, in contrast, offer key advantages as sources of biomarkers. NDEVs are derived from neurons and thus provide a more specific reflection of the pathological state of the nervous system. Their molecular markers are directly related to the development and progression of CNS diseases, and they can be analyzed to detect early disease changes, which aids in early diagnosis. Therefore, changes in miRNA levels in NDEVs hold more promise for PD diagnostics than those in other EVs. It has been shown that, in the stress conditions associated with PD, miR-30a-5p and miR-181c-5p levels in NDEVs are reduced. These miRNAs appear to be neuroprotective, mitigating neuronal death by regulating mitochondrial complex I/V function and apoptosis. However, these observations were primarily obtained in studies using neuroblastoma cell lines such as SK-N-SH and SH-SY5Y (Currim et al., 2024), and further clinical validations in patient cohorts are crucial for confirming these findings. One study demonstrated that miR-128, an miRNA enriched in neurons, was significantly reduced in circulating EVs isolated from plasma samples from PD patients compared to those from healthy people. Functionally, miR-128 has been shown to prevent 6-OHDA-induced neuronal death by preventing the activation of transcription factor FoxO3a and regulating both endogenous and exogenous apoptosis pathways. This altered expression of circulating exosomal miR-128 suggests it has a critical role in PD pathogenesis and holds promise as a biomarker for early disease detection (Bhattacharyya et al., 2022). A marked increase in miR-155 levels was observed in NDEVs from PD patients, suggesting it is involved in modulating α-syn-mediated inflammatory responses and contributes to disease progression (Anastasi et al., 2021). In another study analyzing brain-derived blood exosomal miRNAs from 25 PD patients at different disease stages, six miRNAs—miR-195-5p, miR-495-3p, miR-23b-3p, miR-30c-2-3p, miR-323a-3p, and miR-27a-3p—were significantly upregulated in PD plasma samples compared to controls. Notably, the levels of four miRNAs, miR-195-5p, miR-495-3p, miR-323a-3p, and miR-30b-5p, correlated positively with the severity of Hoehn and Yahr staging, indicating their potential utility for monitoring disease progression (Jang et al., 2024).

While these studies have underscored NDEV miRNAs’ distinct advantages and diagnostic potential, research in this area needs to be completed. In contrast, investigations focusing on general EVs and their miRNA cargos have been more extensive. Given the critical utility of NDEVs in reflecting CNS pathology with high specificity, further explorations of NDEV-specific miRNAs are essential. Such efforts will enhance their value as reliable biomarkers for PD diagnosis and disease monitoring and bridge existing research gaps in the field.

Long non-coding RNAs as biomarkers for Parkinson’s disease

LncRNAs, a class of non-coding RNA molecules with lengths exceeding 200 nucleotides, play significant roles in gene expression and regulation at various levels (Statello et al., 2021; Sharma et al., 2024). LncRNAs in neural-derived L1CAM EVs have been associated with the autophagic-lysosomal pathway, linking them to α-syn production (Zou et al., 2020). Lysosomal dysfunction is implicated in increased EVs secretion, highlighting the significance of lncRNAs in EVs, particularly L1CAM EVs, to PD development. The consistent detectability of exosomal lncRNAs in plasma suggests they have potential utility as biomarkers (Nie et al., 2021; Zhang et al., 2021; Wang et al., 2022). Research revealed that linc-POU3F3 levels were markedly elevated in plasma L1CAM EVs from PD patients relative to those of healthy controls, and the degree of elevation correlated with the severity of PD, suggesting Linc-POU3F3 is a potential biomarker for facilitating the diagnosis of PD and monitoring its progression. Elevated linc-POU3F3 levels may suppress the expression of autophagy-related proteins, leading to impaired α-syn degradation, enhanced α-syn release in EVs and accelerating PD progression (Zou et al., 2020).

Other potential biomarkers for Parkinson’s disease

Lipids and cytokines are both potential PD biomarkers. Lipids are indispensable in the physiological functions of EVs. For example, phosphatidylinositol phosphate (PIP) is a vital lipid group involved in vesicular transport and is capable of responding to various signals that influence the generation and secretion of EVs. LPS-induced type I interferon inhibits PIP-5-kinase-1-gamma expression, resulting in elevated PI4P levels in MVBs and the recruitment of RAB10 and thereby enhancing small EV production. Macrophages treated with LPS for 24 hours showed upregulated expression of HSPA5, a heat shock protein. The interaction between PI4P and HSPA5 in the Golgi apparatus or endoplasmic reticulum, regions distant from MVBs, was shown to hinder the release of EVs. However, exosomal lipid PI4P modulated intraluminal vesicle formation to regulate EV secretion (Jin et al., 2023a).

EVs also facilitate the transport of cytokines. For instance, the cargos of EVs derived from pericytes include various cytokines, such as growth factors, IL-8, IL-6, and MCP-1, which are associated with diverse physiological activities, exemplified by angiogenesis, BBB integrity, and neurotrophic support (Sharma et al., 2022). Research has indicated that growth factors such as glial cell line-derived neurotrophic factor (GDNF)can exert neuroprotective effects, and GDNF therapy has demonstrated efficacy in protecting and restoring dopaminergic neurons in aged primates and rodent models of PD (Kells et al., 2012; Haney et al., 2020). A promising therapeutic strategy involves the release of EVs containing neurotrophic growth factors; these effectively transport GDNF to neurons and thereby enhance the survival of dopaminergic neurons (Zhao et al., 2014; Whone et al., 2019). A study indicated that BDNF levels within plasma EVs remained consistent across PD and healthy individuals. However, the data demonstrated a significant correlation between plasma EV BDNF and motor performance. BDNF levels were markedly decreased in PD patients exhibiting severe dyskinesia compared to individuals with milder symptoms, especially patients with postural instability and gait disturbance (PIGD)-associated symptoms (Chung et al., 2020b). PIGD is considered a more severe subtype of PD, potentially arising from initially benign forms (Lee et al., 2019). Therefore, tracking the levels of BDNF and other cytokines in plasma-derived EVs could potentially be used to detect and monitor the progression of PD. Plasma EVs from PD patients also exhibit increased levels of pro-IL-1β and TNF-α relative to those from healthy controls. The severity of PIGD and cognitive symptoms, as measured via the Mini-Mental State Examination and the Montreal Cognitive Assessment score, correlated significantly with the concentrations of IL-1β, TNF-α, IL-6, and IL-10 in plasma-derived EVs (Chan et al., 2021).

Researchers have identified various other candidate biomarkers within NDEVs in addition to α-syn. EVs play a role in regulating iron metabolism by transporting ferritin and the transferrin receptor (TfR). Remarkably elevated concentrations of ferritin and TfR were discovered in the plasma NDVEs of individuals with PD compared to healthy controls, and TfR was identified as an independent predictor for PD. This suggested that EVs may contribute to the excessive iron deposition within PD (Chen et al., 2023). The increasing trend in NfL levels in PD patients with severe akinetic rigidity suggests that NfL levels in plasma-derived EVs may indicate the severity of motor symptoms (Chung et al., 2020a; Ng et al., 2020). Notably, tau and Aβ_1–42 were not effective in differentiating between PD and healthy individuals, but elevated levels of these proteins in plasma EVs were significantly linked to cognitive dysfunction in PD patients (Chung et al., 2021a). Subsequent studies showed changes in the levels of α-syn, tau, and Aβ_1–42 with disease duration in PD patients. The elevated content of these pathogenic proteins was observed in PD cases with more severe motor and cognitive deficits, indicating they are associated with clinical deterioration (Chan et al., 2023). Thus, tau and Aβ_1–42 have been identified as promising biomarkers for diagnosing PD.

A comprehensive analysis of NDEVs isolated from the plasma of PD patients and age-matched healthy controls revealed significant differences in both the vesicle characteristics and molecular profiles of the two groups. Transmission electron microscopy and subsequent nanoparticle tracking analysis demonstrated that NDEVs from PD patients were larger than those from healthy controls. Proteomic profiling identified 23 key proteins associated with the KEGG pathway in NDEVs from PD patients, including PARK7, gelsolin, and clusterin, which were present at lower levels in healthy people. These findings highlighted the potential of NDEVs as a valuable source of biomarkers for the diagnosis of PD and investigations of its underlying pathological mechanisms (Anastasi et al., 2021; Figure 4).

Figure 4.

Key biomarkers found in BDEVs in peripheral blood.

Protein markers associated with neuronal damage and synaptic dysfunction include α-syn, p-α-syn, DJ-1, tau, and Aβ₄₂. Inflammatory cytokines such as IL-1β, TNF-α, and IL-6 serve as indicators of neuroinflammation. Additionally, RNA molecules such as miRNA, lncRNA, and circRNA are involved in gene regulation. These biomarkers hold potential diagnostic value and may offer insights into the mechanisms underlying disease progression. Created with BioRender.com. Aβ₄₂: Amyloid-beta 42; BDEVs: brain-derived extracellular vesicles; BDNF: brain-derived neurotrophic factor; IL: interleukin; NfL: neurofilament light chain; SNARE: soluble N-ethylmaleimide-sensitive factor attachment protein receptors; Tf: transferrin; TfR: transferrin receptor; TNF-α: tumor necrosis factor-alpha; α-syn: α-synuclein.

Extracellular Vesicles for Parkinson’s Disease Treatment

EVs simultaneously provide highly promising therapeutic delivery systems for diverse applications and can serve as key agents for facilitating the transmission of pathogenic α-syn that can be targeted for therapeutic intervention. Therapeutic strategies focusing on EVs may involve modulating the production of EVs carrying pathogenic α-syn, disrupting the elimination of EVs during transmission, and inhibiting the uptake of EVs by receptor neurons.

Decreasing extracellular vesicles biogenesis

One possible therapeutic direction is to inhibit EV production, as EVs are carriers of pathologic α-syn, and the proteins required for EV formation may serve as therapeutic targets (Gurung et al., 2021). The neutral enzyme sphingomyelinase is necessary for facilitating an ESCRT-independent process of EV biogenesis. GW4869 effectively inhibits neutral sphingomyelinase, thereby obstructing exosome production through the disruption of intraluminal vesicle formation (Catalano and O’Driscoll, 2020). GW4869 has been shown to inhibit the secretion of EVs from activated microglia, particularly those activated by α-syn, and can inhibit the release of EVs containing histone L. Another neutral sphingomyelinase inhibitor, DDL-112, inhibits α-syn aggregation in SN and reduces the biosynthesis of EVs (Zhu et al., 2021). Several inhibitors of EV release, such as calpeptin, D-panthioethylamine, and imipramine, are currently under investigation. Initiating autophagy to disrupt the initial stages of α-syn exosome release and absorption could present an innovative method to curb the progression of PD and related synucleinopathies.

Inhibiting the uptake of extracellular vesicles

Exosomes influence target cells through their phagocytic uptake, membrane fusion, and interactions between receptors and ligands (Mulcahy et al., 2014; Mathieu et al., 2019; van Niel et al., 2022). Endocytosis has been implicated as the predominant pathway for exosome internalization; in this process, the plasma membrane engulfs intact exosomes, leading to their incorporation into the endosomal system (Gonda et al., 2019; Gurung et al., 2021). There is significant value in studying α-syn in microglial-derived EVs, as 4%–12% of exosomes in the CSF of PD patients are of microglia/macrophage origin (Guo et al., 2020). In the CSF, α-syn primarily originates from the CNS, with 2.17% present in CSF exosomes, which is an adequate amount to trigger the aggregation of α-syn and the degeneration of dopaminergic neurons (Mollenhauer et al., 2012; Stuendl et al., 2016). The α-syn in exosomes more readily enters other neurons, inducing α-syn aggregation, while free α-syn is distributed on the cell surface (Delenclos et al., 2017; Xia et al., 2021). A study revealed that, after EVs from the plasma of PD patients were injected into the striatum of mice, microglia took up a substantial quantity of the EVs, and numerous α-syn oligomers were detected on the EVs’ exteriors (Xia et al., 2021). Moreover, TLR2 binds to oligomeric α-syn and promotes the internalization of α-syn by microglia (Kim et al., 2013; Song et al., 2021). Aging and the development of neurodegenerative diseases make microglia less efficient in removing proteins. These undegraded proteins interfere with proteasome and lysosome functions, further promoting PD development (Choi et al., 2015; Hoenen et al., 2016). Therefore, inhibiting TLR2 could potentially hinder the propagation of α-syn and have a therapeutic effect on PD (Kim et al., 2018). Possible directions for removing pathogenic EVs from circulation or preventing pathogenic EVs from entering recipient cells should be explored.

Extracellular vesicles as therapeutic delivery systems

A number of therapeutic delivery platforms, including ligand-modified nanoparticles, micelles, and dendrimers, are currently being researched. However, exosomes offer several advantages over these systems. First, they are derived from human cells, leading to their low immunogenicity and high biocompatibility (Ha et al., 2016; Meng et al., 2020). Second, exosomes possess a membrane structure identical to cell membranes, which aids target cells internalize their contents. Third, exosomes are nanosized, preventing their rapid elimination through the mononuclear phagocyte system (Van Den Boorn et al., 2011; Ridolfi and Abdel-Haq, 2018). Fourthly, exosomes and their contents exist stably and long-term in the blood circulatory system, from where they can modulate the peripheral immune system and transfer brain antigens to the periphery (De Rivero Vaccari et al., 2016; Elliott and He, 2021). This ability enables exosomes to cross various biological barriers. Lastly, exosomes in the blood exhibit significant neurological targeting capabilities and can directly cross the BBB without modification (Alvarez-Erviti et al., 2011b; Qu et al., 2018). In summary, exosomes represent a research hotspot for the development of therapeutic programs.

Engineered extracellular vesicles

Naturally secreted exosomes exhibit limited capabilities for drug uptake. However, engineered exosome mimics, which combine synthetic liposomes or nanoparticles with functional components of natural exosomes, have broader applications as drug carriers. Genetically engineered mammalian cells equipped with devices for exosomal transfer have been developed to enable the efficient production of designer exosomes. These exosomes feature enhanced mRNA packaging and improved delivery capabilities, demonstrating therapeutic potential for PD. Researchers used this system to deliver catalase mRNA in the PD model, causing a decline in neuroinflammation and neurotoxicity (Kojima et al., 2018). Another team explored strategies for loading high-abundance functional mRNAs into exosomes, which yielded positive outcomes (Yang et al., 2019). Technologies, such as composite exosomes, exosome-mimetic nanovesicles, and synthetic exosomes, are currently being developed for diverse applications (Mondal et al., 2023). Immature dendritic-cell-derived exosomes have been shown to be highly suitable for modification, as the absence of immune-associated molecules on the surface of immature dendritic cells, including CD40 and MHC-II, reduces their immunogenicity and virulence (Heras-Murillo et al., 2024). A team validated the feasibility of using engineered EVs in treating PD. Initially, exosomes were modified to display the rabies virus glycoprotein (RVG) peptide on their surface and encapsulate anti-α-syn short hairpin RNAs (shRNA) minicircles. These RVG-EVs were intravenously injected into a mouse model, leading to a sustained decrease in α-syn levels within the spinal cord, distal intestine, and brain (Izco et al., 2019; Izco et al., 2023). Further research showed that RVG-EVs can specifically target the nervous system after intravenous injection, avoiding drug delivery to the heart or liver and bypassing immune system activation and making them suitable for long-term therapy. Thus, the use of anti-α-syn shRNA-MC RVG-EVs represents a promising therapeutic approach (Izco et al., 2023).

Drug-loading methods

Two primary strategies for incorporating cargo into exosomes are available: exogenous and endogenous loading techniques (Sutaria et al., 2017). The exogenous pathway involves isolating exosomes before drugs are loaded via processes such as active and passive loading (Ingato et al., 2016). Passive loading occurs when the drug is incubated with exosomes, while active loading involves disrupting the exosome membrane. In endogenous loading, exosomes encapsulate the drugs before their secretion.

Electroporation is a crucial method for exogenous drug loading and is particularly suitable for siRNA and miRNA cargo (Shtam et al., 2013; Tian et al., 2014). However, whether the loaded RNA retains its functionality is a subject that requires further investigation (Lamichhane et al., 2015). Saponin is commonly used to permeabilize exosome membranes, facilitating the incorporation of catalase into macrophage-derived exosomes, which mitigates oxidative stress and protects neurons in PD (Haney et al., 2015). A third approach involves hypotonic dialysis, which induces osmotic changes to facilitate the drug loading of exosomes (Tan et al., 2015).

An endogenous loading method is transfection, which allows the loading of oligonucleotides into exosomes. GDNF, expressed by glial cells, alleviates neurodegeneration in PD patients. Studies indicate that EVs derived from macrophages exhibit an enhanced ability to target inflamed brain tissues compared to those from neuronal or glial cells. Genetically modified macrophages have been used to produce GDNF-carrying EVs by introducing a GDNF-encoding plasmid DNA into parental cells. These exosomes demonstrated prolonged efficacy in Parkin Q311 (X)A transgenic mice for reducing brain inflammation and enhancing motor function (Zhao et al., 2022).

Extracellular vesicles carry a variety of therapeutic substances

EVs transport therapeutic substances such as proteins, miRNAs, and siRNAs, which are crucial in PD development and show potential as therapeutic drugs. Exosomes have been shown to be effective carriers for siRNA. RVG exosomes with α-syn siRNA were delivered to normal and S129D α-syn transgenic mice, which significantly reduced α-syn levels in the brain after 1 week. This resulted in a marked reduction in protein aggregates within neurons and protection of dopaminergic neurons in the SN. SNCA-targeted siRNAs also specifically downregulate α-syn expression (Cooper et al., 2014). RVG-modified exosomes can transport siRNAs from the circulatory system into the CNS, reducing α-syn levels in the brain and suppressing the mRNA and protein expression of BACE1 in mice (Alvarez-Erviti et al., 2011b). However, the relatively short half-life of siRNAs has prompted exploration into the use of shRNAs for longer-term gene silencing. RVG exosomes effectively deliver shRNA-MC constructs to the brain, decreasing α-syn expression and aggregation (Izco et al., 2019). Curcumin, known for its neuroprotective properties and ability to reduce existing α-syn, possesses anti-inflammatory and neuroprotective properties (Singh et al., 2013; Sharma and Nehru, 2018). Curcumin-containing exosomes have demonstrated efficacy in resisting lipopolysaccharide-induced neuroinflammation in a mouse model (Singh et al., 2013). Compared with directly delivered curcumin, exosome-loaded curcumin results in a threefold increase in anti-inflammatory activity (Zhuang et al., 2011). Combining siSNCA and curcumin is a promising approach for treating PD. However, owing to low bioavailability and challenges in crossing the BBB, delivering both drugs via modified exosomes has emerged as a viable therapeutic strategy.

In PD pathogenesis, patients exhibit reduced brain antioxidant enzymes, compromising the control of neuroinflammation and neuronal damage. Exosomes can transport catalase across the BBB, acting on neurons and microglia to reduce oxidative stress-induced damage in the brain (Haney et al., 2015). Antisense oligonucleotides have been demonstrated to effectively suppress α-syn expression. The exosome-mediated delivery of antisense oligonucleotides4 markedly mitigated α-syn aggregation induced by α-syn PFFs. Mice exhibit suppressed α-syn expression within the brain, leading to amelioration of dopaminergic neuron degeneration and improved motor function (Yang et al., 2021). Activated microglia impair autophagy and exhibit high expression of PELI1, an E3 ubiquitin ligase (Xiao et al., 2013; Du et al., 2017). The specific mechanism involves PFFs acting on microglia, triggering cellular activation. Elevated PELI1 expression disrupts lysosomal function, inhibiting autophagy and increasing exosomal secretion, which promotes α-syn metastasis (Baixauli et al., 2014; Guo et al., 2020). Silencing endogenous PELI1 expression via siRNA is a viable approach, facilitating autophagy recovery and reducing α-syn aggregation. In summary, the use of EVs as drug carriers has great potential for managing PD.

Therapeutic potential of extracellular vesicles in dopaminergic neuron protection and repair

EVs hold significant potential as mediators of dopaminergic neuroprotection and regeneration, with glial cells playing diverse roles in influencing these processes. It has been shown that miR-34a in astrocytic shedding vesicles is markedly upregulated following LPS stimulation. When transferred to dopaminergic neurons, miR-34a downregulates the antiapoptotic protein Bcl-2, reducing neuronal resilience to neurotoxic insults. Inhibition of astrocytic miR-34a expression has been shown to restore the antiapoptotic capacity of dopaminergic neurons in vitro and attenuate 6-OHDA-induced neuronal loss and motor impairments in vivo (Mao et al., 2015). Recent studies have demonstrated the neuroprotective potential of platelet-derived EVs (PEVs) from human platelet concentrate supernatants as a promising therapeutic approach for PD. PEVs carry neurotrophic factors, anti-inflammatory proteins, and antioxidants, enabling multitargeted protective effects. In vitro experiments revealed that PEVs restore dopaminergic neuron function and mitigate ferroptosis-induced damage. Moreover, in vivo studies have confirmed their ability to cross the BBB via intranasal administration. This results in the protection of tyrosine hydroxylase-positive neurons in the substantia nigra and the alleviation of motor deficits in PD mouse models (Delila et al., 2024). PEVs exert their effects through the modulation of neuroinflammation by regulating microglial activity, thus suppressing proinflammatory cytokine release. Molecular analyses revealed that PEVs are enriched with BDNF, platelet-derived growth factor, and transforming growth factor β1, in addition to antioxidants such as glutathione peroxidase, superoxide dismutase 1, and superoxide dismutase 2. These molecules act synergistically to protect dopaminergic neurons by inhibiting lipid peroxidation and ferroptosis, key drivers of neurodegeneration (Zhu et al., 2009; Goldie et al., 2014; Lu et al., 2019). Additionally, PEVs enhance neuronal differentiation and maturation, as demonstrated by increased β-III tubulin expression in SH-SY5Y cells, which supports neuronal functional recovery. Owing to their neuroprotective, anti-inflammatory, and neuroregenerative properties, combined with their ability to cross the BBB and favorable safety profile, PEVs represent promising, noninvasive therapeutic options for PD. Their multimodal effects offer significant clinical potential for improving the treatment and management of this debilitating neurodegenerative disease. In BV-2 microglial cell models, PEVs attenuate LPS-induced inflammatory responses by decreasing the protein levels of proinflammatory cytokines such as TNF-α, IL-6, and IL-1β, thereby improving the neuronal microenvironment (Xie et al., 2018).

Stem cell-derived EVs have demonstrated considerable promise in protecting and regenerating dopaminergic neurons. Human nasal mucosal mesenchymal stem cells (hnmMSCs) are considered an ideal source for dopaminergic neuron differentiation because of their accessibility, ease of cultivation, and robust differentiation capacity (Alizadeh et al., 2019). During the differentiation of dopaminergic progenitor cells, small EVs (da-hnmMSC-sEVs) are secreted, which can effectively cross the BBB and target manganese-induced neural injury sites when delivered intranasally. These vesicles significantly alleviate neuroinflammation and promote the functional recovery of neurons. Mechanistically, Da-hnmMSC-sEVs carry miR-494-3p, which downregulates CMPK2 and NLRP3 expression. This pathway supports neurogenesis and provides substantial protection to dopaminergic neurons, highlighting da-hnmMSC-sEVs as promising therapeutic candidates for neuroinflammatory and neurodegenerative diseases, including PD (Yang et al., 2024). The clinical translation of EVs remains limited by challenges in achieving efficient brain delivery and crossing the BBB.

A promising solution lies in the development of dopamine-conjugated EVs, which are engineered by conjugating dopamine to the surface of adipose-derived stem cell-derived EVs. These dopamine-conjugated EVs selectively target dopaminergic neurons, enhancing their cellular uptake and significantly increasing their brain accumulation in PD models. In PD mouse models, intravenous administration of dopamine-conjugated EVs improved brain distribution and motor function recovery and reduced dopaminergic neurodegeneration. Mechanistically, these therapeutic effects are linked to autophagic pathway activation, as evidenced by the upregulation of the Beclin-1 and LC3-II proteins. This innovative EV engineering strategy offers a targeted, practical approach for BBB penetration, highlighting its potential as a therapeutic platform for PD treatment (Sul et al., 2024; Figure 5).

Figure 5.

EVs as drug carriers for the treatment of Parkinson’s disease.

Therapeutic substances, such as drugs, proteins, siRNAs, and other molecules, can be loaded into EVs using either endogenous or exogenous methods. Exogenous loading techniques include passive incubation and active methods such as electroporation, saponin treatment, freeze-thaw cycling, sonication, and extrusion. EVs have low immunogenicity and can cross the blood–brain barrier, which increases the bioavailability of the loaded drug while reducing its toxicity. Targeted EVs can be engineered through artificial modification, allowing for the delivery of therapeutic molecules to the brain after intravenous injection. Created with BioRender.com. EVs: Extracellular vesicles; ESCRT: endsomal sorting complex required for transport.

Roles of brain-derived extracellular vesicles in neural regeneration

Recent studies have highlighted the therapeutic potential of BDEVs in promoting neural regeneration. EVs derived from neurons, astrocytes, microglia, and Schwann cells represent distinct subtypes of BDEVs, each playing a unique role in neural communication, regeneration, and repair. These vesicles have been shown to carry specific cargo, such as neurotrophic factors, miRNAs, and other bioactive molecules. Leveraging the unique properties of these BDEVs in therapeutic strategies may provide innovative approaches for enhancing neuroregeneration and addressing the progressive neurodegeneration in PD.

Neuron-derived extracellular vesicles in neural regeneration

NDEVs have emerged as key mediators of nerve regeneration by modulating the injury microenvironment and facilitating stem cell differentiation into neuron-like cells. Coculture studies have demonstrated that small neuronal EVs can drive the differentiation of adipose-derived stem cells into neuron-like cells. This process is regulated primarily by essential molecules in NDEVs, including the synaptic-related protein SNAP25 and miRNAs such as miR-132 and miR-9. In sciatic nerve injury animal models, adipose-derived stem cells differentiated under the influence of NDEVs have been shown to improve nerve regeneration, reduce neuronal degeneration, and restore axonal density (Roballo et al., 2019).

In spinal cord injury, NDEVs aid in neural repair by remodeling the injured microenvironment. In the acute phase, NDEVs prevent overactivation of astrocyte and microglia activity, reducing the inflammatory response and the formation of glial scars. In addition, NDEVs may promote the differentiation of neural stem cells into functional neurons (Xu et al., 2023). Notably, under inflammatory conditions, NDEVs upregulate the expression of miR-21, which suppresses SMAD7 and activates the TGF-β/SMAD2 signaling pathway. Amplification of this signaling pathway not only promotes the differentiation of neural stem cells into astrocytes but also exacerbates glial scar formation, resulting in the inhibition of axonal regeneration and remyelination. While this pathway supports NSC differentiation into astrocytes, it can also exacerbate glial scar formation, hindering axonal regeneration and remyelination. NDEVs play multiple roles in nerve regeneration, but more in-depth research is needed (Han et al., 2023).

Microglia-derived extracellular vesicles in neural regeneration

MDEVs are critical mediators of neural regeneration, primarily through their ability to regulate inflammation and improve the neural microenvironment. Emerging evidence suggests that EVs secreted by polarized M2 microglia can effectively reduce neuroinflammation and enhance neurological recovery. For example, M2 microglial EVs are enriched with specific miRNAs, such as miR-7670-3p, which regulate the expression of ATF6, a key factor involved in endoplasmic reticulum stress (Chen et al., 2023a). By suppressing the production of proinflammatory mediators and reducing neurotoxicity, these vesicles help stabilize inflammatory conditions in neurological disorders. Furthermore, in AD model mice, MDEVs attenuate amyloid-β accumulation, which helps to protect the integrity of neuronal dendritic spines and improve synaptic function. These vesicles carry regeneration-associated factors that enhance myelin sheath repair, optimize the microenvironment of injured nerves, and support axonal regeneration. Importantly, MDEVs can regulate the endoplasmic reticulum stress response and promote neuronal survival by translocating miRNAs (Casella et al., 2018; Lombardi et al., 2019).

Astrocyte-derived extracellular vesicles in neural regeneration

ADEVs are integral to neuroprotection and nerve regeneration because of their bioactive cargo. ADEVs are enriched with growth factors such as fibroblast growth factor-2 and vascular endothelial growth factor, which play pivotal roles in neural repair. Fibroblast growth factor-2 promotes neural stem cell proliferation, migration, and differentiation while facilitating axonal growth and regeneration by activating fibroblast growth factor receptors and the mitogen-activated protein kinase signaling pathway. It enhances hippocampal neurogenesis and confers neuroprotection, as demonstrated in rodent stroke models via intracerebroventricular infusion and nanoliposome-mediated delivery (Pearson et al., 2001; Rai et al., 2007; Zhao et al., 2016). Vascular endothelial growth factor complements these effects by driving cerebral angiogenesis, synaptic plasticity, and neurogenesis, accelerating the restoration of neural networks and improving functional recovery (Tillo et al., 2012).

In addition to growth factors, ADEVs contain HSP70, a cytoprotective molecule that stabilizes misfolded proteins and mitigates cellular stress (Venediktov et al., 2023). HSP70 exerts neuroprotective effects by inhibiting caspase activation, blocking apoptotic vesicle formation, and preventing neuronal apoptosis (Vinokurov et al., 2024).

Furthermore, ADEVs are rich in stress-inducible protein 1, which interacts with the cell surface protein PrPc. This interaction triggers the ERK1/2 signaling pathway, promoting neuronal survival and differentiation. Stress-inducible protein 1 also supports neuroprotection through the cAMP-dependent protein kinase A pathway, enhancing neuronal resilience under stress conditions (Roffé et al., 2010). ADEVs play multifaceted roles in neuroprotection and nerve regeneration by delivering critical growth factors, cytoprotective proteins, and signaling molecules.

Schwann cell-derived extracellular vesicles in neural regeneration

Schwann cell-derived EVs (SC-EVs) play a vital role in nerve repair and regeneration. Following peripheral nerve injury, Schwann cells undergo dedifferentiation, adopting a reparative phenotype that releases EVs enriched with bioactive molecules to support the injury site. SC-EVs facilitate axonal regeneration, enhance neuronal survival, and modulate the neural microenvironment through intercellular signaling.

In an optic nerve injury model, SC-EVs were internalized by retinal ganglion cells, where they significantly improved cell survival and stimulated axonal regeneration and elongation via activation of the cAMP/CREB signaling pathway. Additionally, SC-EVs mitigate glial overactivation and reduce neurodegeneration associated with optic nerve damage (Zhu et al., 2023). Similarly, in sciatic nerve injury models, SC-EVs promoted axonal growth by regulating growth cone dynamics and inhibiting the GTPase activity of RhoA, a known suppressor of axonal regeneration (Lopez-Verrilli et al., 2013). SC-EVs have also shown therapeutic potential in CNS injuries. In spinal cord injury, SC-EVs reduce the deposition of chondroitin sulfate proteoglycans, major inhibitors of axonal regeneration, by modulating the expression of Toll-like receptor two in astrocytes. Concurrently, they promote neuronal survival and motor function recovery by activating the NF-κB/PI3K signaling pathway (Pan et al., 2021). At the molecular level, SC-EVs are enriched with factors critical for nerve regeneration. GAP43 promotes axon elongation and growth cone guidance through cytoskeletal remodeling, whereas tau proteins stabilize microtubules to support axonal integrity. RAC1 further facilitates precise axonal regeneration by regulating actin dynamics. Additionally, SC-EVs carry functional microRNAs, such as those targeting TIMP3 and PTEN, which activate the PI3K/AKT signaling pathway to reduce neuronal apoptosis and accelerate neuronal growth (Hua et al., 2015; Zhou et al., 2015).

In summary, BDEVs, including those released by neurons, glial cells, and Schwann cells, are essential mediators of neuroprotection, tissue repair, and intercellular communication within the CNS. The degeneration and loss of dopaminergic neurons represent the core pathological hallmark of PD, positioning the protection and regeneration of these neurons as a critical focus in therapeutic research. Studies exploring neural regeneration across various neurological disorders have highlighted the potential of BDEVs to facilitate neural repair and functional recovery. These findings underscore the importance of incorporating BDEVs into treatment strategies for PD, particularly in the context of promoting dopaminergic neuron regeneration (Figure 6).

Figure 6.

Involvement of BDEVs in neurodegeneration.

Neuron-derived EVs facilitate the differentiation of stem cells into neuron-like cells and play a critical role in regulating the neuroinflammatory environment by inhibiting the overactivation of microglia and astrocytes. MDEVs promote neuroprotection by alleviating endoplasmic reticulum stress and reducing inflammation-induced damage. Additionally, MDEVs have been shown to enhance myelin repair and promote axonal regeneration. Astrocyte-derived EVs deliver bioactive molecules, including growth factors and heat shock protein 70, which stimulate axonal regeneration and suppress neuronal apoptosis. Schwann cell-derived EVs enhance both axonal regeneration and elongation while providing neuroprotective effects. Collectively, these BDEVs exhibit diverse roles in facilitating neural repair and regeneration. Created with BioRender.com. BDEVs: Brain-derived EVs; EVs: extracellular vesicles; MDEVs: microglia-derived EVs.

Limitations

This review highlights the potential of BDEVs in PD research while acknowledging key limitations. Current studies on BDEVs, particularly NDEVs, predominantly focus on α-syn as a biomarker, with limited investigations into other molecular cargos, such as miRNAs and lncRNAs. Although the prevailing view is that L1CAM can serve as a marker for NDEVs, its specificity remains controversial. In addition, few studies have directly compared BDEVs with EVs, so the unique role of BDEVs in PD remains to be further defined. In addition, although the results of preclinical studies show promise for BDEVs as potential therapeutic vectors, large-scale clinical studies have not yet been conducted. Overcoming these challenges will require standardized methodologies, rigorous comparative analyses, and robust clinical validation to realize the full potential of BDEVs in the diagnosis and treatment of PD.

Discussion

BDEVs have demonstrated outstanding potential for Parkinson’s mechanism research, clinical diagnosis, and treatment. However, current research still has many limitations.

Research on the role of BDEVs in PD pathogenesis is still in its early stages. While evidence suggests that BDEVs contribute to the spread of pathogenic proteins such as α-syn and influence neuroinflammation and oxidative stress, the precise molecular and cellular mechanisms underlying these processes are not fully understood. Further investigations are needed to clarify how BDEVs facilitate neurodegeneration and disease progression in patients with PD.

In terms of diagnostics, while α-syn in BDEVs has been well explored, there is limited research on miRNAs as potential diagnostic biomarkers. Current studies often analyze total EVs from body fluids rather than specifically isolating BDEVs, which diminishes the specificity of miRNA-based diagnostics. It is essential to differentiate between general EV populations and BDEVs to accurately identify biomarkers for PD diagnosis, disease monitoring, and disease progression.

In addition, studies comparing BDEVs with general EVs are highly limited, and direct comparisons are necessary to further highlight the unique advantages of BDEVs in the diagnosis and treatment of PD. Functionally, BDEVs may play unique roles in synaptic remodeling, neuroinflammatory regulation, and the propagation of α-syn aggregation.

Comparative studies on these functional differences could validate the specific contributions of BDEVs to neuroprotection and neuroregeneration in PD. Additionally, the efficiency of BDEVs versus general EVs in crossing the BBB remains an open question. Investigating the molecular mechanisms underlying BBB permeability, including potential receptor-mediated pathways, could provide crucial insights into the translational applications of BDEVs. Systematic comparisons between BDEVs and general EVs in PD are essential for understanding their biomarker specificity, functional roles, and capacity to cross the BBB.

The therapeutic application of BDEVs in PD remains in its early stages, with only a limited number of studies exploring their potential. A significant barrier is the absence of standardized BDEV isolation and characterization methods, particularly for neuron-derived EVs. L1CAM is currently the most widely used marker for identifying NDEVs. However, concerns about its specificity have emerged. L1CAM is not restricted to neurons but is also expressed in CNS cells, such as oligodendrocytes, and peripheral nervous system cells, including Schwann cells. Additionally, its presence has been detected in various nonneuronal cell types, such as monocytes, T cells, B cells, endothelial cells, and certain cancers. Notably, the cross-reactivity of the L1CAM antibody with α-syn further complicates its reliability, potentially confounding α-syn detection in EV-based assays. Although isolation and purification techniques are still underdeveloped, BDEVs have shown outstanding potential in protecting and repairing dopaminergic neurons. Continued efforts to establish reliable protocols for biomarker detection are critical for translating research into viable diagnostic and therapeutic applications for PD.

Future research needs to focus more on the following aspects. First, standardized protocols for the isolation and identification of BDEVs must be established to improve the credibility and reproducibility of research results. Given the limitations and ongoing debate regarding L1CAM as a neuronal marker, future studies should explore alternative markers or marker combinations to improve the accuracy of NDEV isolation. Second, although α-syn in NDEVs has been extensively studied as a biomarker in multiple clinical cohorts, clinical studies of other biomarkers in NDEVs remain very limited. Efforts should be focused on identifying and validating additional potential biomarkers in NDEVs. Comparative studies analyzing NDEVs against total EV populations will help clarify the unique biological and pathological roles that BDEVs play in PD progression. Third, while the connection between EVs and PD is undeniable, the exact extent of the involvement of EVs in PD development remains unclear and requires quantitative studies. The autonomous selection of cargoes by EVs and the regulatory mechanisms of their signaling processes are still unknown. Leveraging advanced tools, such as in vivo models, will provide deeper insights into the pathways governed by BDEVs, offering a more comprehensive understanding of their contributions to neurodegeneration and disease progression. Finally, the therapeutic potential of BDEVs demands rigorous preclinical validation. Engineered BDEVs, which are designed to deliver neuroprotective molecules, such as neurotrophic factors, antioxidants, or gene-modulating tools, represent promising strategies to increase dopaminergic neuron survival, reduce neuroinflammation, and stimulate neurogenesis. Optimizing delivery methods, particularly noninvasive approaches such as intranasal administration, is critical to ensure efficient BBB penetration with minimal off-target effects. Translational studies assessing the safety, optimal dosage, and therapeutic efficacy of BDEVs in PD-relevant models are essential for bridging the gap in clinical application.

These advances have the potential to drive the creation of new minimally invasive, targeted, and precise clinical diagnostic and treatment protocols to provide better care for patients with PD.

Conclusions

In summary, BDEVs can serve as potential indicators of CNS responses and biomarkers for PD diagnosis, reflecting the status of and changes in originating cells. BDEVs crucially contribute to PD onset and progression by facilitating intercellular communication. Moreover, BDEVs can be detected in the blood because of their capacity to traverse the BBB, and exogenous EVs can function as therapeutic drug carriers that target the brain. Diverse components of EVs have been recognized as biomarkers for distinguishing PD. EVs represent a therapeutic target and an ideal drug delivery system for PD, offering extensive opportunities for therapeutic research. BDEVs also offer broad research prospects in the PD field. Further research is vital to elucidate the association between BDEVs and PD, enabling BDEVs to play a more significant role in the mechanistic research, diagnosis, and treatment of PD.

Funding Statement

Funding: This work was supported by the National Natural Science Foundation of China, No. 82271278; 2019 Wuhan Huanghe Talents Program; 2020 Wuhan Medical Research Project, No. 2020020601012303; 2021 Hubei Youth Top-notch Talent Training Program and 2022 Outstanding Youth Project of Natural Science Foundation of Hubei Province, No. 2022CFA106; and Medical Research Program of Huatongguokang, No. 2023HT036 (all to NX).

Data availability statement:

Not applicable.