Parkinson's disease (PD) is the second most common neurodegenerative disorder, with only partial symptomatic therapy and no disease modifying therapies. The accumulation and aggregation of the protein alpha-synuclein (αsyn) is causatively linked both genetically and pathologically to PD (Polymeropoulos et al., 1997; Spillantini et al., 1997). αSyn is a small presynaptic protein generally found under physiological conditions as a monomeric, unfolded, and soluble protein. However, several factors can trigger αsyn misfolding leading to the formation of intermediate soluble species, oligomers, and protofibrils, which become insoluble β-sheet-rich fibrils and are found deposited into the cytosol of affected neurons and are known as Lewy bodies (Forno, 1996; Spillantini et al., 1997; Halliday et al., 2011). A large body of evidence points to the prefibrillar αsyn oligomers as a source of αsyn-induced toxicity (Conway et al., 2000; Volles et al., 2001; Lashuel et al., 2002; Sharon et al., 2003; Kayed et al., 2004; Danzer et al., 2007, 2011; Winner et al., 2011) and although the exact mechanism of αsyn-induced toxicity remains unknown, inhibition of αsyn oligomerization has been explored as a potential therapeutic strategy and explains the growing interest in monitoring these species. Another feature of PD is the possible spread of misfolded αsyn along distinct neuroanatomical pathways suggesting a sequential transfer and misfolding event underlies the pathology and progression of the disease. The propagation of aggregated proteins between cells has been proposed to be an important mechanism of disease propagation not only in PD (El-Agnaf et al., 2006; Desplats et al., 2009; Olanow and Prusiner, 2009) but also other neurodegenerative disorders such as Alzheimer's disease, Amyotrophic lateral sclerosis (ALS), or Huntington's disease (Ren et al., 2009; Münch et al., 2011; Guo and Lee, 2014). The process by which oligomers are internalized into target cells and exit cells remains largely unknown and has resulted in great debate in the literature. The involvement of extracellular vesicles (EVs) such as exosomes has received increasing attention in synucleinopathies and in the field of neurodegeneration in general (Ghidoni et al., 2008; De Toro et al., 2015; Quek and Hill, 2016). EVs can function as shuttles for the delivery of cargo between cells and play a crucial role in cell-to-cell communication. Interestingly, in the last decade a number of proteins involved in neurodegenerative disorders have been identified in their normal and pathogenic states to be associated with EVs and in particular exosomes including PrP, tau, Aβ, SOD1, and most importantly αsyn (Rajendran et al., 2006; Gomes et al., 2007; Vella et al., 2007; Danzer et al., 2012; Saman et al., 2012). This discovery underscores a potential pathogenic role for EVs as vehicles to transfer toxic aggregated proteins between cells in disease conditions.

Exosomes are small membrane vesicles ranging from 40 to 100 nm in diameter that represent one population of EVs. Various cell types including neurons, astrocytes, and glia, have the capacity to release these vesicles into the extracellular space where they are later found in body fluids such as urine, blood, and cerebrospinal fluid (CSF; Théry et al., 2006; Keller et al., 2011). Exosomes originate from the inward budding of endosomes into multivesicular bodies (MVBs) to form intraluminal vesicles that are then directed outside of the cells (Kowal et al., 2014). They can carry lipids, proteins, and nucleic acids that are eventually released into neighboring cells cytoplasm. The process of internalization is still elusive but can occur through direct fusion of the plasma membrane or required receptor-mediated endocytosis. Depending on the recipient cells, internalization has been described to occur through various mechanisms such as clathrin-mediated endocytosis in neuronal cells (Frühbeis et al., 2013; Tian et al., 2014), clathrin-independent but cholesterol- and lipid raft-dependent endocytosis in endothelial and some tumor cells (Svensson et al., 2013) or caveolae-dependent endocytosis in epithelial cells (Nanbo et al., 2013).

EVs seem to play an important role in PD transmission and consequently have generated great interest for potential applications in both diagnostics and therapeutics of neurodegenerative disorders. Thus, it appears crucial to better understand their molecular mechanisms of interaction with target cells. The objective of the present study is to elucidate the role of exosomes in the transmission of oligomeric forms of αsyn and shed new light on the route(s) leading to their internalization into recipient cells. We previously described a cell based assay where exosome–associated αsyn oligomers could be produced and monitored using a highly sensitive bioluminescence protein complementation assay (Danzer et al., 2012). Taking advantage of this new technology, we purified exosomes from conditioned media of a newly generated stable cells line (Moussaud et al., 2015) secreting αsyn oligomers and investigated endocytic pathway(s) required for their uptake. An understanding of these events will clarify the therapeutic potential of enzymes that regulate protein trafficking and degradation in synucleinopathies.

EVs, and in particular exosomes, appear to play an important role in several physiological and pathological processes. They are seen as delivery machines capable of traveling between cells and unloading their contents across cell membranes mediating inter-cellular transfer. Several cell types can release these nanoparticles in the extracellular space and a number of aggregated proteins involved in neurodegenerative disease have been associated with EVs. A body of evidence suggests that exosomes may be involved in the spreading of misfolded neurodegenerative disease-associated proteins such as Prp, Tau, Aβ, and αsyn, and may be efficient initiators of disease propagation (Bellingham et al., 2012; Schneider and Simons, 2013). Recent studies have shown that exosomal αsyn can be detected in human cerebrospinal fluids (CSF; Stuendl et al., 2016) and blood plasma from PD patients (Shi et al., 2014) further emphasizing the importance of these nanovesicles in αsyn spreading mechanism. However, the exosome–cell interaction mode and their intracellular trafficking pathway in recipient cells remains incompletely understood. In this study, exosome internalization associated with αsyn oligomers was examined in a cellular assay. As previously reported by our group we confirmed oligomeric species of αsyn are detected in both the exosomal pellet and the exosome-free supernatant from the conditioned media (Danzer et al., 2012). Using an in vitro uptake assay combined with luciferase activity assay we provide direct evidence that free-αsyn oligomers cannot be internalized efficiently into recipient cells but only associate with the external surface of plasma membranes, as they could be removed by simple trypsinization. In contrast, trypsinization had no effect on exosome-treated cells because exosomes were readily internalized inside recipient cells. Our observations imply that although oligomeric αsyn occur in the extracellular milieu independent of exosomes, a small fraction of extracellular αsyn is released within EVs and is more prone internalization by recipient cells.

One of the central issues associated with secreted αsyn is to understand which species of αsyn are internalized by recipient cells that will eventually initiate intracellular toxic effects. In our hands, secreted vesicular αsyn appears to be readily internalized by neuroglioma cells compared to free αsyn oligomers and we and others have previously shown that exosomal αsyn confers toxicity on neighboring cells (Emmanouilidou et al., 2010; Danzer et al., 2012). By utilizing a protein complementation assay we directly track soluble αsyn oligomers and monitor αsyn aggregation with a surrogate luciferase readout. Therefore, we cannot exclude the possibility that free monomeric αsyn or free high molecular weight species of αsyn are efficiently internalized by neuroglioma recipient cells. Although in a parallel experiment (not shown) we performed an ELISA assay and still failed to detect internalization of αsyn in recipient cells indicating that free monomeric species were not taken up. In support of our data recent evidence suggests that αsyn uptake by neuronal cells (Luk et al., 2009) depends on the fibrilarization state of αsyn and cationic liposomes have been shown to be necessary for the internalization of recombinant fibrillar αsyn species into neuronal cells. Furthermore, Lee et al. (2008) suggested that aggregated αsyn is internalized more efficiently and via a different pathway, compared to monomeric αsyn protein. They suggest that monomeric αsyn directly translocate through the plasma membrane, which we were unable to measure with our complementation assay. It also appears that oligomers of αsyn exert greater cytotoxicity in recipient cells than soluble monomers (Desplats et al., 2009; Emmanouilidou et al., 2010), however, these data should be interpreted with caution, because they are based on high amounts of recombinant protein. Of note, our experimental setup used only cell-produced αsyn oligomers.

As described earlier, more free αsyn oligomers are detected in the preparation than exosome-associated. Although, the origin of free αsyn oligomers is still unclear it could be explained by extracellular degradation of the exosomal membrane by proteases or lipases allowing the release of proteins from the exosomal lumen to the extracellular matrix (Hughes, 1999). Nevertheless, in support of our observation only a small fraction of Aβ peptide is found associated with exosomes (Rajendran et al., 2006), and EV-associated proteins have been found in amyloid plaques of post mortem human brains (Thompson et al., 2016). In addition, Grey et al. showed that the propensity of αsyn to aggregate is increased by the presence of exosomes (Grey et al., 2015). Thus, one could speculate that exosomes may provide a favorable environment for the oligomerization process. In fact it has been shown that lipid-mediated oligomerization seems to be important in amyloid formation and polyunsaturated fatty acids have been shown to trigger multimerization of recombinant αsyn (Perrin et al., 2001). Interestingly when exosomes were isolated from CSF of PD patients, Stuendl et al. reported that the EVs were able to induce αsyn aggregation in a naïve cell line (Stuendl et al., 2016). The authors speculate that pathogenic αsyn oligomers may be preferentially sorted into exosomes and act as a seed into recipient cells. It is tempting to believe that exosomes may carry a pathogenic from of αsyn however, further in depth studies will be needed to determine if this is the case.

Since cell-to-cell propagation of pathogenic proteins has been deemed important in PD and neurodegenerative disorders, it is of interest to understand the pathway by which exosomes enter into cells. Most experimental evidence suggests that EVs are usually taken up via endocytosis (Mulcahy et al., 2014) but yet there appears to be little agreement in the literature as to which type of endocytic mechanisms are most important, with clathrin-dependent, caveolae dependent (Svensson et al., 2013), macropinocytosis (Fitzner et al., 2011; Tian et al., 2014), phagocytosis lipid raft-mediated uptake (Feng et al., 2010) or direct fusion with plasma membrane being the current postulates. Exosomal uptake is extremely rapid. We identified exosomes inside cells as early as 30 min after exposure. Our data also demonstrate that incubation at 4°C significantly reduces αsyn oligomer internalization, suggesting that uptake is an energy-requiring process as previously described (Christianson et al., 2013; Mulcahy et al., 2014; Tian et al., 2014). However, 30% of our exosomal population still gained access into the cells despite the low temperature. We therefore cannot neglect the fact that direct fusion with the plasma membrane may occur. Also, macropinocytosis-, clathrin-, and caveolin-dependent endocytosis inhibitors had no significant effect on exosome internalization in our assay. In a control experiment (not shown), exosomes isolated from cells in the presence of tetracycline and therefore not overexpressing αsyn underwent internalization that could not be blocked by cytochalasin D or nystatin suggesting that the presence of αsyn oligomers in exosomes does not contribute to the lack of inhibition. These differences might reflect the heterogeneity both in EV populations and in the cell types being used in the literature. The context of experiments may also affect the outcome and account for the observed discrepancies and different entry routes might reflect cell specialization or conditions. We can also speculate that multiple entry routes might even coexist in the same cell. Indeed any of the common endocytosis inhibitors used in our study abrogate exosome internalization suggesting that they may simultaneously trigger a number of different gateways and that different mechanisms are involved in internalization of recipient cell. Further experiments with tools such as antibodies to prevent receptor ligand interactions or RNAi may be of help to elucidate the endocytosis mechanisms involved.

Lastly various complexes, including viral particles and lipoproteins, use HSPGs to help gain entry into cells. Recently, cancer cell exosomes have been shown to depend on cell-surface HSPGs for their internalization (Christianson et al., 2013; Franzen et al., 2014). Interestingly, treatment of EVs with heparinase to remove surface proteoglycans had no effect on uptake, suggesting that it is the presence of HSPGs on the cell surface that are important for mediating vesicular entry (Christianson et al., 2013). Furthermore, Holmes et al. (2013) suggested that HSPGs mediate the uptake and seeding of αsyn recombinant fibrils. Even though evidence tends to suggest a crucial role of HSPGs in internalization of exosomes we could not observe this effect in our assay as cells genetically modified to be depleted of proteoglycans were still able to internalize exosome-associated αsyn. It is important to note that in the Holmes et al. study (Holmes et al., 2013) recombinant fibrillar proteins were used, which may explain the different results. Also, this result strongly correlates with our hypothesis that exosomes may gain entry into a cell via more than one route.

Further advancement of our understanding of both the EV uptake mechanism and interaction of αsyn with EVs are crucial and will enable development of therapeutic strategies for PD and others neurodegenerative disorders. Preventing the release of exosomes may become a novel and attractive approach to stop the spreading of oligomers of αsyn.

MD carried out the study design, analysis, and drafting of the manuscript. TT isolated the exosomes and carried out the cellular uptake assay. AB and DM assisted with cell maintenance and isolation of the exosomes. SM provided the stable cell line and participated in the drafting and editing of the manuscript. IY and TP carried out the nanosight reading. PM participated in the experimental design, coordination, interpretation, drafting and editing of the manuscript. All authors read and approved the final manuscript.