. Cell-produced alpha-synuclein is secreted in a calcium-dependent manner by exosomes and impacts neuronal survival. J Neurosci. 2010 May 19;30(20):6838-51. PubMed.


Please login to recommend the paper.


  1. The ability of many cell types, both prokaryotic and eukaryotic, to disseminate and retrieve biological material is increasingly apparent. The purpose of such exchange in many instances remains unclear, and in the case of shared pathogenic protein aggregates, even seems counterproductive. Is one cell’s trash another’s (Trojan) treasure? Depending on the mechanism, this exchange involves varying levels of specificity, and an effective but relatively non-specific means that is beginning to garner needed attention in neurodegenerative diseases is via exosomes, tiny vesicles formed from the endocytosis of a small segment of invaginated cell membrane, which are eventually released into the extracellular space. The ability of exosomes to transport numerous macromolecules over long distances suggests that they could serve as vectors for the prion-like spread of proteopathies (Aguzzi and Rajendran, 2009).

    Emmanouilidou and colleagues, in a comprehensive set of experiments, provide evidence for a role or exosomes in the spread of cytotoxic forms of cellularly generated α-synuclein from cell to cell in vitro. The secretion of vesicles is calcium-dependent, and the cytotoxicity is mitigated by immunodepletion of α-synuclein or interference with oligomers. The findings support the idea that Parkinson’s-type synucleinopathy can be transferred among cells (e.g., Desplats et al., 2009; Luk et al., 2009; Kordower and Brundin, 2009). The uptake of secreted α-synuclein by cycling SH-SY5Y cells, but not (non-cycling) neurons, remains an important issue for additional work, as the authors note, inasmuch as the state of multimerization of the protein could be a factor. However, this nice study, in the context of recent work on other disorders, reinforces the view (Miller, 2009) that the prion-like seeding of protein aggregation is a common feature of several neurodegenerative diseases, including Alzheimer disease. The potential role of exosomes in this process clearly deserves further exploration.


    . The transcellular spread of cytosolic amyloids, prions, and prionoids. Neuron. 2009 Dec 24;64(6):783-90. PubMed.

    . Inclusion formation and neuronal cell death through neuron-to-neuron transmission of alpha-synuclein. Proc Natl Acad Sci U S A. 2009 Aug 4;106(31):13010-5. PubMed.

    . Exogenous alpha-synuclein fibrils seed the formation of Lewy body-like intracellular inclusions in cultured cells. Proc Natl Acad Sci U S A. 2009 Nov 24;106(47):20051-6. PubMed.

    . Propagation of host disease to grafted neurons: accumulating evidence. Exp Neurol. 2009 Dec;220(2):224-5. PubMed.

    . Neurodegeneration. Could they all be prion diseases?. Science. 2009 Dec 4;326(5958):1337-9. PubMed.

  2. Cytosolic Amyloids: Being Out Is In
    In the last few months, the neurodegeneration community has witnessed a paradigm shift in the way we understand the spread of amyloids in the brain. Several reports suggested a prion-like behavior of amyloid proteins such as α-synuclein, tau, and huntingtin. [Editor’s note: see ARF Live Discussion.] These amyloids indeed seem to be released from cells and then effect the conversion of their monomeric counterparts in the neighboring cells/grafts. At the same time, there are two major reasons why these amyloids are fundamentally different from prions. First, prions are transmissible between humans/animals; second, they are confined to the lumenal side of the cell, whereas α-synuclein, tau, and huntingtin amyloids are cytoplasmic in nature. Therefore, a puzzling question arises: how do these amyloids get released from the cell and re-enter the neighboring cell (or the target graft as in the case of the Parkinson’s stem cell transplants)?

    One could envision several possibilities. Amyloids might acquire a property of interacting with membrane lipids to mediate their translocation; they might leach out of a part of the membrane, or make transient pores and get released out of the cell. There is no convincing evidence for any such mechanism. Non-classical secretion of cytokines and growth factors has been observed, for example, fibroblast growth factor 2, IL-1β, annexins, migration inhibitory factor1, galectins, but at the moment we do not understand the precise mechanism behind this phenomenon. Several possibilities exist (Aguzzi and Rajendran, 2009), one mechanism being through the exosomal pathway. This paper from the group of Kostas Vekrellis in Athens now suggests that α-synuclein takes this route.

    First author Evangelia Emmanouilidou and colleagues used an inducible cell system to study the controlled release of α-synuclein. When induced, these cells produced and released α-synuclein, a finding that confirms previous observations (Desplats et al., 2009; Luk et al., 2009). [Editor’s note: see also ARF SfN story.] But Vekrellis’s group asked: how? Through a series of simple but necessary cell-based experiments, they establish that:

    1. α-synuclein doesn’t pass the lipid bilayer;
    2. overexpression was not the reason for the release;
    3. serum factors shown to affect non-classical secretion of proteins did affect α-synuclein release; and
    4. heat shock increased the release of α-synuclein.

    To understand the mechanisms by which α-synuclein is externalized, the Athens group set out a series of rigorous cell biological experiments. They show that neither scaffolding (actin) nor signaling proteins (protein kinases) played a major role in the release. Instead, they found two familiar players to be crucially involved in the exocytosis of α-synuclein. Increasing cytosolic calcium boosted release of α-synuclein, as did blocking acidification of endocytic compartments. Since both endocytosis and calcium have been previously linked to the release of exosomes—small vesicles that are released from the cells through the endocytic pathway—the authors suspected (rightfully so!) that perhaps α-synuclein utilizes the exosomal pathway to get released via this route.

    It is worth mentioning that Seung-Jae Lee and colleagues have previously demonstrated that interfering with lysosomal acidification indeed affected the release of α-synuclein (Desplats et al., 2009), but now Vekrellis and colleagues show that lysosomal acidification affects α-synuclein release presumably via release of exosomes (also see Vingtdeux et al., 2007).

    How do exosomes release cytosolic proteins such as α-synuclein? We know more about endocytic mechanisms. They have largely been associated with internalization and release of lumenal contents (i.e., coming from the extracellular space) and membrane proteins. What’s less well known is that the exosomal pathway also releases a considerable number of cytosolic proteins into the extracellular space. During endocytosis, plasma membrane invagination (outside-in) gives rise to early endosomes. Their limiting membrane undergoes another round of invagination (which is now inside-out) to form the intralumenal vesicles, giving the endosome a multivesicular appearance. During this invagination, these intralumenal vesicles encapsulate a significant amount of cytosol with them. Multivesicular bodies harboring such intralumenal vesicles can now fuse with the plasma membrane to release these ILVs as exosomes. This also explains the topology of exosomes being identical to that of the plasma membrane (outside-out; inside-in) with the cytosol encapsulated within them (Aguzzi and Rajendran, 2009).

    The current work from Vekrellis’s group shows that this is indeed the case. Through a battery of biochemical experiments, the authors show that α-synuclein is released via exosomes. Since α-synuclein could associate with lipid membranes, the authors looked at whether α-synuclein in exosomal vesicles is membrane-associated (to the inner leaflet of the vesicle) or in the soluble part, and they found that it is associated with both fractions. This is an interesting finding because lipid-mediated oligomerization of α-synuclein has been found to be important for its toxicity and the amyloid conformation (also in other amyloids: see Wang et al., 2010 on lipids and PrP, and Yuyama et al., 2008; Yuyama and Yanagisawa, 2009).

    Next, the Greek authors looked at whether the released α-synuclein is toxic to cells. Conditioned media from α-synuclein expressing cells induced toxicity that was rescued when pre-incubated with α-synuclein antibody. Though both monomeric and oligomeric α-synuclein were found in the extracellular medium, this toxicity seems, at least partially, to be mediated by α-synuclein oligomers (of the high- and low-molecular-weight kind). Proliferating cells, not differentiated cells, were particularly vulnerable to the toxicity. Perhaps remodeling of the cell membrane during cell division makes cells particularly vulnerable for oligomer-induced toxicity. Moreover, compounds that interfered with oligomerization inhibited the toxicity.

    In summary, this study is important in many aspects. For the first time, it elegantly demonstrates that cytosolic amyloids can be released on exosomes, and it suggests a cellular mechanism for their release. I believe exosomal release could be a key mechanism for the spread of amyloids. That may well have implications for cell-based therapy. It would be interesting to know if exosome-associated α-synuclein was indeed the agent that was responsible for the reported seeding effect in the grafted tissues (Brundin et al., 2008). Perhaps blocking exosome release could become one way to inhibit the spread. Currently, there are no means or tools available that selectively inhibit the release of exosomes.

    As is true with any good study, certain limitations exist and new questions emerge:

    1. How and why was α-synuclein toxic to the cells? Is the membrane-associated α-synuclein or the free soluble α-synuclein the culprit? In all likelihood, it was the soluble oligomeric α-synuclein. That’s because immunodepletion of α-synuclein from the conditioned media reduced the toxicity and exosome-encapsulated α-synuclein would be inaccessible to the antibody. Then, what is the role of exosome-associated α-synuclein?

    2. How is soluble, non-exosomal α-synuclein secreted? This seems to parallel the Aβ case. We showed that even though Aβ is generated by β/γ-secretase-mediated cleavage of APP in early endosomes and released via the exosomal pathway (Rajendran et al., 2006), actually only a small fraction of this Aβ is associated with exosomes. Several explanations exist for this puzzle. For example, Aβ peptides might become unstable and lose their affinity to lipid membranes after being released from the cell, due to differences in pH or ionic concentrations. Thereby, they might contribute to the soluble pool. Perhaps this could explain the soluble fraction of the released α-synuclein, as well.

    3. Beyond this intriguing finding, the presence of oligomers is worth pondering. While the authors do not show if oligomers are associated with exosomal vesicles, lipid-mediated oligomerization seems to be an important issue in amyloid formation. In fact, high-resolution lipidomic analysis of exosomes shows enrichment of lipids involved in amyloid formation (our unpublished data). Thus, exosomes could mediate on one hand the release of cytosolic α-synuclein and, on the other, provide an environment that is conducive for oligomerization processes.

    4. This study merely showed that α-synuclein is released via exosomes. It did not show specifically if exosome-associated α-synuclein is toxic to the cells, nor did it show if exosome-associated α-synuclein is then taken up by neighboring cells to mediate the seeding effect. Perhaps it is only the soluble fraction that is toxic. In a way, that would be a good thing. It would imply that diffusion-limited spread represents a main cause for the seeding, not a vesicle-mediated process. Besides the fact that these amyloids are entirely different from prions in that they are incapable of transmission from person to person, this study suggests that there is no danger that long-range transmission by these cytosolic amyloids could occur.


    . The transcellular spread of cytosolic amyloids, prions, and prionoids. Neuron. 2009 Dec 24;64(6):783-90. PubMed.

    . Inclusion formation and neuronal cell death through neuron-to-neuron transmission of alpha-synuclein. Proc Natl Acad Sci U S A. 2009 Aug 4;106(31):13010-5. PubMed.

    . Exogenous alpha-synuclein fibrils seed the formation of Lewy body-like intracellular inclusions in cultured cells. Proc Natl Acad Sci U S A. 2009 Nov 24;106(47):20051-6. PubMed.

    . Alkalizing drugs induce accumulation of amyloid precursor protein by-products in luminal vesicles of multivesicular bodies. J Biol Chem. 2007 Jun 22;282(25):18197-205. PubMed.

    . Generating a prion with bacterially expressed recombinant prion protein. Science. 2010 Feb 26;327(5969):1132-5. PubMed.

    . Accelerated release of exosome-associated GM1 ganglioside (GM1) by endocytic pathway abnormality: another putative pathway for GM1-induced amyloid fibril formation. J Neurochem. 2008 Apr;105(1):217-24. PubMed.

    . Late endocytic dysfunction as a putative cause of amyloid fibril formation in Alzheimer's disease. J Neurochem. 2009 Jun;109(5):1250-60. PubMed.

    . Research in motion: the enigma of Parkinson's disease pathology spread. Nat Rev Neurosci. 2008 Oct;9(10):741-5. PubMed.

    . Alzheimer's disease beta-amyloid peptides are released in association with exosomes. Proc Natl Acad Sci U S A. 2006 Jul 25;103(30):11172-7. PubMed.

  3. Extracellular α-Synuclein: Multiple roles for the same protein

    Without doubt the role of secreted α-synuclein needs to be characterized further. Our data suggests that synuclein may be exerting its effects extracellularly either by entering proliferating cells or acting solely on the cell membrane as is the case with neurons. Whether these effects are mediated via a still-unidentified receptor remains to be examined. We failed to observe synuclein internalization by neuronal cells; however, we cannot rule out the possibility that specific oligomeric species may be internalized by neuronal cells but are too minute in amount to be detected by our labeling assay.

    Our study further points toward “free” and exosome-associated alpha-synuclein having different roles in the extracellular space. However, in our study we did not attempt to establish a toxic role for exosome-associated synuclein. This is indeed a question that remains to be answered, especially in light of the observed increase of secreted synuclein levels after treatment of our cells with acidotropic agents that affect the endocytic pathway. In this respect it would be also interesting to examine how (Macro) autophagy may influence the synthesis and release of exosomes.

    It is also possible that exosomes could be a mechanism of confinement and removal of the toxic intracellular synuclein species; as such exosomal release of synuclein could be a way of cell defense against the toxic agent. Our data that synuclein presumably “toxic” oligomers are readily detectable in exosomes is in favor of this idea. In this respect, the involvement of microglia in this clearance would also be worth investigating. Alternatively, exosomes may be a way of modifying synuclein (since exosomes–as all other vesicular compartments-are thought to have a specific environment).

  4. The same has been previously described with respect to tau protein. Thus, extracellular tau protein is toxic for SH-SY5Y (Gomez-Ramos et al., 2006). Aggregated and phosphorylated tau are less toxic than dephosphorylated tau. In addition, tau increases intracellular calcium likely through muscarinic receptors (Gomez-Ramos et al., 2008, 2009). Thus, the extracellular toxicity of tau protein, and now α-synuclein, suggest a common mechanism to explain propagation in those diseases.


    . Extracellular tau is toxic to neuronal cells. FEBS Lett. 2006 Sep 4;580(20):4842-50. PubMed.

    . Characteristics and consequences of muscarinic receptor activation by tau protein. Eur Neuropsychopharmacol. 2009 Oct;19(10):708-17. PubMed.

    . Extracellular tau promotes intracellular calcium increase through M1 and M3 muscarinic receptors in neuronal cells. Mol Cell Neurosci. 2008 Apr;37(4):673-81. PubMed.

  5. I'd like to submit a technical question for clarification. I was very excited to see this paper. But when I saw the sequence in Figure 3E, I was surprised, because although the sequence looked familiar, two amino acids seemed out of place. To make sure that I was not remembering the sequence wrong, I performed a blast search using the published peptide and found that this peptide is from β-synuclein.

    Here are the sequences of the two tryptic peptides:

    EGVV_Q_GVA_S_VAEK is β-synuclein
    EGVV_h_GVA_t_VAEK for α-synuclein

    β-synuclein sequence is: link

    α-synuclein sequence is: link

    This is a little troubling, since it cast into question in my mind if the protein that was transfected into the cells was actually α-synuclein? The α and β isoforms are largely similar, so I would expect the antibodies used in this study to pick up both isoforms. In addition, the size difference on an SDS page gel would be small enough that it would be difficult to distinguish the α and β isoforms based on their size. Would the authors kindly want to address this point?

  6. Reply to comment by Tim West
    I am happy to clarify this question. First, I would like to point out that the antibodies used for the detection of α-synuclein in our cell-system are specific to α-synuclein (see also Vekrellis et al., 2009). Indeed, the correct sequence for the α-synuclein tryptic peptide under question is:


    From this study, a total of two peptides were detected that collectively corroborate the α-synuclein identification.

    The tandem mass spectrum shown in our publication was chosen on the basis of a better signal-to-noise ratio. However, this tandem mass spectrum suggests a Glu>pyro-Glu modification at the N-terminus and exhibits a low peptide sequence coverage. The additional tandem mass spectrum detected in this study translated to the amino acid sequence (-)TKEQVTNVGGAVVTGVTAVAQK(-) (observed with one miscleavage at 95 percent ID confidence in concordance to the Mascot software and validated with the Scaffold software program and further verified with manual de novo sequencing interpretation) is uniquely surrogate to α-synuclein. Collectively, the proteomics evidence strongly favors the α-synuclein protein identification over that of β-synuclein.

    The annotated tandem mass spectral evidence for: (-)TKEQVTNVGGAVVTGVTAVAQK(-) with over 80 percent peptide sequence coverage is as follows:

    (A) Observed and interpreted tandem mass spectrum (see PDF A).

    (B) Tabulation of the observed product B (red) and Y (blue) ions from theabove spectrum and concordant amino acid sequence interpretation (see PDF B).


    . Inducible over-expression of wild type alpha-synuclein in human neuronal cells leads to caspase-dependent non-apoptotic death. J Neurochem. 2009 Jun;109(5):1348-62. PubMed.

  7. In their discussion, Emmanouilidou et al. consider the possibility that the degenerative effects of extracellular aggregates of α-synuclein on differentiated SH-SY5Y cells and primary cortical neurons are mediated by a specific receptor or by the formation of membrane pores. The neurotrophin receptor p75 is a suitable candidate for such a receptor. According to evidence provided in the Aβ-crosslinker-hypothesis [.pdf], aggregates of NAC, a natural fragment of α-synuclein, can activate p75 and induce neurite budding and apoptosis via p75, and these effects can be prevented by administration of a juxtamembrane stalk fragment of p75 that is part of the stalk binding site of Aβ on p75. The hypothesis argues that Aβ, which is known to interact with α-synuclein, crosslinks p75 with α-synuclein species and thereby mediates certain protective and deleterious effects of p75 and α-synuclein.

Make a Comment

To make a comment you must login or register.