The Transcellular Propagation and Intracellular Trafficking of α-Synuclein

Abstract

Parkinson’s disease is the second most common neurodegenerative disorder, with only partial symptomatic therapy and no mechanism-based therapies. The accumulation and aggregation of α-synuclein is causatively linked to the sporadic form of the disease, which accounts for 95% of cases. The pathology is a result of a gain of toxic function of misfolded α-synuclein conformers, which can template the aggregation of soluble monomers and lead to cellular dysfunction, at least partly by interfering with membrane fusion events at synaptic terminals. Here, we discuss the transcellular propagation and intracellular trafficking of α-synuclein and posit that endosomal processing could be a point of convergence between these two routes. Understanding these events will clarify the therapeutic potential of enzymes that regulate protein trafficking and degradation in synucleinopathies.

Aggregated α-synuclein can spread from cell to cell by uptake and release, templating further aggregation that leads to disease (e.g., Parkinson’s). Endosomal processing may serve as a point of convergence for these processes.

Parkinson’s disease (PD) is the second most common neurodegenerative disorder, affecting 1% of people over the age of 60. Clinically, it is characterized primarily by a movement disorder causing resting tremor, bradykinesia, rigidity, postural instability, and diverse nonmotor symptoms including dementia, which in community-based studies, was reported in up to 80% of patients with a long disease duration (Hely et al. 2008). This latter finding indicates that PD is a diffuse neurodegenerative disorder. Similarly, detailed neuropathological studies have shown that one of the cardinal histological features, the intraneuronal inclusions called Lewy pathology, is detected in numerous cortical areas and often correlates with the extent of cognitive decline (Spillantini et al. 1997; Goedert et al. 2013). Despite this diffuse evolution, the presentation to health services is commonly a result of the loss of a critical number of dopaminergic neurons in the substantia nigra (Lees et al. 2009), but in patients with dementia with Lewy bodies (DLB), dementia may be the presenting feature.

The identification of causative genes in familial cases and their study in animal models has suggested new intracellular and cell-non-autonomous mechanisms of pathogenicity. A major challenge is the prioritization and validation of those mechanisms that can best explain the histological characteristics of the disease (i.e., the intraneuronal accumulation of α-synuclein in Lewy pathology, which is a sine qua non feature of the most common form of sporadic PD and detected postmortem in >95% of cases). α-Synuclein is also the major component of glial cytoplasmic inclusions that define multiple system atrophy (MSA), a movement disorder characterized by cerebellar ataxia, parkinsonism, and autonomic dysfunction. In MSA, unlike PD and DLB, many α-synuclein inclusions are found in glial cells (Spillantini et al. 1998a,b).

Therefore, understanding the molecular mechanisms that underpin the pathogenicity of α-synuclein is important for the development of therapies that aim to benefit the vast majority of patients with PD, DLB, and MSA. Besides its presence in the form of insoluble inclusions in all cases of disease (Goedert 2015), the pathogenicity of α-synuclein is also supported by the following: (1) cases of PD and DLB are caused by missense mutations or multiplications in the α-synuclein gene (SNCA) (Polymeropoulos et al. 1997; Goedert 2015); (2) polymorphisms in SNCA are the most common risk factor for idiopathic PD and MSA (Goedert 2015); and (3) overexpression of α-synuclein is associated with aggregation and neuronal toxicity in model organisms (Feany and Bender 2000; Osterberg et al. 2015). Mechanistic studies in vitro and in vivo suggest that aggregated α-synuclein is toxic through a gain-of-function mechanism (reviewed in Tofaris and Spillantini 2007). This gain of toxic function is caused by α-synuclein misfolding, which acts as a template for further aggregation and leads to cellular dysfunction, at least partly, by interfering with fusion events in vesicle recycling (Perrett et al. 2015).

In this review, we discuss the transcellular propagation and intracellular trafficking of α-synuclein and posit that endosomal processing could be a point of convergence between these routes (Fig. 1).

Figure 1.

Proposed model for the propagation and intracellular trafficking of α-synuclein at the endosome. α-Synuclein fibrils bind to surface receptors on neuronal plasma membrane such as LAG3 and are internalized in endosomes. Receptor levels per se may be regulated by ubiqutination and endosomal processing. Intracellular α-synuclein is targeted to endosomes (e.g., by Nedd4-dependent ubiquitination), where it may encounter internalized fibrils. The acidic pH of endosomes would be a permissive environment for template-induced misfolding of monomers and amplification of fibrillar conformers. Such conformers may be degraded by lysosomes or under certain conditions (e.g., by endosomal rupture) released in the cytosol where they could act as seeds for further aggregation of cytosolic α-synuclein monomers. EE, Early endosome; LE, late endosome; E3, ubiquitin ligase; Ub, ubiquitin.

TRANSCELLULAR PROPAGATION OF α-SYNUCLEIN

α-Synuclein is enriched in presynaptic nerve terminals and loosely associates with membranes through its N-terminal part. This interaction with lipids induces its folding into α-helical conformers (Burré et al. 2014). In cells, upon cross-linking, α-synuclein can be detected as metastable oligomers, predominantly tetramers, which are in equilibrium with monomers (Dettmer et al. 2015). In solution, monomeric α-synuclein is natively unfolded with a propensity to self-assemble into β-sheet-rich filaments, with the same ultrastructure as those extracted from Lewy pathology (Spillantini et al. 1998a,b). In vitro, the conversion of soluble α-synuclein into amyloid fibrils typically occurs after a lag phase that is followed by a rapid increase in fibril formation and is concentration-dependent (Li et al. 2001). This suggests that a critical amount of α-synuclein in amyloid precursors may stochastically form in solution, acting as “seeds” to promote the recruitment of unfolded α-synuclein in amyloidogenic permissive structures and the synergistic formation of amyloid fibrils. In addition, α-synuclein was shown to assemble into two different strains with different structures and seeding properties (Bousset et al. 2013). Therefore, at least in vitro, α-synuclein possesses properties that could explain its pathogenicity by a conformational templating mechanism. Accordingly, the A53T, H50Q, and E46K α-synuclein mutations have consistently been shown to increase the rate of self-aggregation (Conway et al. 1998; Giasson et al. 2001; Choi et al. 2004; Greenbaum et al. 2005; Ghosh et al. 2013). On the other hand, not all mutations share this property; for example, the A30P, G51D, and A53E mutations reduce the rate of self-aggregation, and also impair the ability of α-synuclein to bind to lipids and brain vesicles (Jensen et al. 1998; Fares et al. 2014; Ghosh et al. 2014).

The concept of pathological templating gained impetus by the finding of Lewy pathology in embryonic neural grafts 11–16 years after transplantation into the brains of people with PD (Kordower et al. 2008; Li et al. 2008). Together, these histological and biochemical studies raised the possibility that misfolded α-synuclein from the host PD brain passed into the grafted cells and templated the conversion of soluble α-synuclein similar to infectious cases of human prion diseases. These studies reinvigorated an earlier suggestion by Braak et al. (2003), whose postmortem neuropathological studies have indicated that in the majority of cases, Lewy pathology evolves along interconnected brain networks that begin in the dorsal motor nucleus of the glossopharyngeal and vagal nerves or the olfactory bulb and the anterior olfactory nucleus, and spread rostrally. It is noteworthy that Lewy pathology is also found in the enteric nervous system, raising the possibility that spread of pathology may also occur between the gastrointestinal system and the brain. In support of gut-to-brain spreading, vagotomy has been associated with a reduced risk of PD (Gray et al. 2015), whereas in animal models, the gut-to-brain evolution of α-synuclein pathology has been reported following the intragastric infusion of rotenone (Pan-Montojo et al. 2010).

In transgenic (Tg) mice, human α-synuclein can transit to nerve cells grafted into the hippocampus (Desplats et al. 2009). Moreover, intracerebral inoculation of brain tissue from symptomatic mice accelerated α-synuclein aggregation in the brains of presymptomatic Tg mice, leading to an earlier onset of neurological symptoms (Mougenot et al. 2012). Unlike induced Aβ and tau assemblies, α-synuclein inclusions in this model were associated with neurodegeneration. Evidence for assembled α-synuclein behaving like a prion has come from the injection of MSA brain extracts into heterozygous mice Tg for A53T human α-synuclein (Watts et al. 2013; Prusiner et al. 2015). Intracerebral injection led to the death of the mice and the development of abundant α-synuclein inclusions. Unlike MSA, where α-synuclein inclusions are mainly found in glial cells, the induced inclusions were present in nerve cells. Injection of recombinant α-synuclein assemblies into the hind limb muscle of Tg mice also induced cerebral α-synuclein aggregation and a neuronal phenotype that was mitigated by transection of the sciatic nerve (Sacino et al. 2013). Spreading of α-synuclein aggregates from the periphery to the brain has also been shown in wild-type (WT) rats (Holmqvist et al. 2014; Peelaerts et al. 2015), but it remains unknown whether these aggregates can seed aggregation of endogenous α-synuclein.

Intrastriatal injection of recombinant mouse α-synuclein assemblies into WT mice gave rise to α-synuclein inclusions and some brain dysfunction (Luk et al. 2012). The same fibrils induced the formation of α-synuclein inclusions in primary cortical neurons from WT mice. Intranigral injection of α-synuclein assemblies into WT mice and monkeys resulted in the formation of inclusions of endogenous α-synuclein (Masuda-Suzukake et al. 2013; Recasens et al. 2014). Based on ultrastructure, two forms of misfolded α-synuclein species have been described (Bousset et al. 2013; Peelaerts et al. 2015) that appear to have distinct pathogenic properties: ribbons and fibrils. When injected into the substantia nigra of the rat, ribbons gave rise to Lewy pathology, whereas fibrils, which did not seed Lewy pathology, led to a loss of dopaminergic neurons. It remains to be seen whether ribbons and fibrils of α-synuclein have their counterparts in human diseases.

INTRACELLULAR TRAFFICKING OF α-SYNUCLEIN

Many studies have shown that monomeric and oligomeric forms of α-synuclein can be taken up and released from various cell types, including neurons, and α-synuclein has been detected in extracellular fluids, such as plasma and cerebrospinal fluid (El-Agnaf et al. 2003). When overexpressed, α-synuclein is also released in exosomes (Emmanouilidou et al. 2010), which are derived from multivesicular bodies. In culture, α-synuclein is secreted by neurons (Lee et al. 2005), and its secretion is increased following stress, including oxidation and inhibition of lysosomal function (Jang et al. 2010). Interestingly, under such conditions, there is increased translocation of α-synuclein into vesicles and release of a predominantly oligomeric form (Desplats et al. 2009; Jang et al. 2010; Danzer et al. 2011). These studies suggest that structurally abnormal or damaged α-synuclein may be shuttled into vesicles and released from neuronal cells via exocytosis. It is now established that fibrillar and oligomeric α-synuclein can also be internalized by endocytosis (reviewed in Lee et al. 2014). Addition of preformed fibrils (PFFs) to primary neurons can induce the formation of intracellular inclusions, which are stained by antibodies specific for α-synuclein phosphorylated at serine 129 (Volpicelli-Daley et al. 2011). Using microfluidic devices to separate the soma and axonal projections from separate but adjacent and synapsed primary neuronal populations, it has been shown that, following uptake, PFFs can move bidirectionally along axons, but synaptic contacts are apparently not required for interneuronal transfer (Volpicelli-Daley et al. 2011; Freundt et al. 2012). It is, therefore, possible that α-synuclein aggregates produced in one neuron can be transmitted to a neighboring cell by exocytosis and endocytosis. Accordingly, disruption of the endosomal GTPase Rab5a and pharmacological inhibition of endocytosis reduced the uptake of α-synuclein (Sung et al. 2001; Hansen et al. 2011). In addition, changes in the expression of Rab11, a GTPase of the recycling endosome, promoted the secretion of α-synuclein and decreased its aggregation and toxicity in Drosophila (Breda et al. 2015). The relevance of endosomal processing in the propagation of proteopathic assemblies is supported by a recent study, which showed that binding of α-synuclein PFFs to the immunoglobulin surface protein lymphocyte-activation gene 3 (LAG3) followed by endosomal internalization, is a critical uptake mechanism in neurons (Mao et al. 2016). However, multiple uptake mechanisms may be at play since binding to heparan sulfate proteoglycans have also been shown to mediate the internalization of aggregated α-synuclein via macropinocytosis (Holmes et al. 2013).

One potential mechanism by which intracellular α-synuclein aggregates may perturb cells is through the disruption of membrane function (Shrivastava et al. 2015), permeability (Lashuel et al. 2002), or fusion, especially in the endosomal and vesicular pathways. The latter mechanism has been extensively investigated in yeast cells, which are heavily dependent on endocytic and secretory pathways (Outeiro and Lindquist 2003). Overexpression of α-synuclein in this model system results in dose-dependent toxicity, accumulation of vacuoles, and aggregation with several Rab GTPases (Gitler et al. 2008; Soper et al. 2008, 2011). In addition, α-synuclein interacts with prenylated Rab acceptor protein (PRA1) (Lee et al. 2011) found in the Golgi and late endosomes and regulates the cycling of Rab GTPases during exocytosis and endocytosis (Abdul-Ghani et al. 2001). Co-transfection of α-synuclein and PRA1 caused defects in vesicle trafficking, possibly by inhibition of Rab recycling (Lee et al. 2011). Fusion of late endosomes with lysosomes or autophagosomes requires the SNARE complex (Nichols and Pelham 1998), which is also important for synaptic vesicle fusion. α-Synuclein has been reported to bind to the SNARE protein synaptobrevin-2 (VAMP2) and is required for SNARE complex assembly at the synapse (Burré et al. 2010, 2014). Endogenous α-synuclein was protective against neurodegeneration caused by deletion of the SNARE chaperone cysteine-string protein α (CSPα) (Chandra et al. 2005), whereas its overexpression in Tg mouse models caused redistribution of SNARE proteins (Garcia-Reitböck et al. 2010; Lim et al. 2011). Addition of PFFs to primary neuronal cultures triggered the accumulation of endogenous α-synuclein in axons and impaired Rab7-positive endosomal transport and fusion with lysosomes (Volpicelli-Daley et al. 2014).

Given that aggregated α-synuclein is readily taken up by a variety of cells, including neurons, and can seed aggregation of expressed α-synuclein, it is likely that defense mechanisms exist that prevent the accumulation of α-synuclein and/or rapidly target misfolded α-synuclein for destruction. Although α-synuclein is degraded by both proteasomes and lysosomes, the latter have emerged as the most relevant degradative pathway in PD pathomechanisms (reviewed in Tofaris 2012). Delivery to lysosomes occurs by chaperone-mediated autophagy, macroautophagy, or the endosomal pathway. Autophagosomes also fuse with late endosomes to form amphisomes, indicating that these two routes to the lysosome are interconnected. Macroautophagy is the least selective process of lysosomal degradation, whereas chaperone-mediated autophagy is based on the recognition of a specific amino acid sequence (KFERQ), a motif found in nearly 30% of cytoplasmic proteins (Cuervo et al. 2004), by the heat shock cognate protein 70, triggering the translocation of the substrate through LAMP2A inside the lysosome (Nixon 2013). On the other hand, conjugation of a specific type of ubiquitin chain to protein substrates is a highly regulated process, which mediates trafficking of selective protein cargoes to lysosomes via the endosomal or autophagic pathway. This posttranslational modification occurs in a three-step catalytic process, involving a ubiquitin-activating enzyme E1, a ubiquitin-conjugating enzyme E2, and a ubiquitin ligase E3. At the proteasome and during endosomal uptake, ubiquitin chains are disassembled by de-ubiquitinating enzymes so that the ubiquitin molecules can be reused in subsequent rounds of degradation, but this action of de-ubiquitinases can also serve to prevent the degradation of substrates. Correct delivery of individual protein substrates or protein complexes to lysosomes typically involves the conjugation of a polyubiquitin chain linked via lysine-63 (K63) or multiple monoubiquitins. This in turn triggers the assembly of a highly conserved machinery, the endosomal complex required for transport (ESCRT), which captures the ubiquitin conjugates on the endosomal membrane. ESCRT complexes comprise four distinct assemblies (ESCRT 0, I, II, or III), which recognize the cargo, associate with the endosomal membrane, and sort protein substrates in intraluminal vesicles (Raiborg and Stenmark 2009). Under certain conditions, which are not fully understood, K63-linked ubiquitin chains are also recognized by the autophagy receptors TAX1BP1, NDP52, NBR1, p62 (SQSTM1), and optineurin, which recruit LC3-coated phagophores to mediate a more selective autophagy.

Both macroautophagy (Webb et al. 2003) and chaperone-mediated autophagy (Cuervo et al. 2004) have been implicated in the lysosomal clearance of α-synuclein and the activation of autophagy in vivo was protective against α-synuclein overexpression (Decressac et al. 2013; Xilouri et al. 2013). α-Synuclein is also found within endosomes (Mak et al. 2010; Hasegawa et al. 2011; Boassa et al. 2013). Endosomal α-synuclein is either targeted for degradation by lysosomes or enters the recycling endosome and is released in a process involving Rab11a and Hsp90 (Liu et al. 2009; Hasegawa et al. 2011). Once delivered to the lysosome, α-synuclein is degraded primarily by cathepsin D (Sevlever et al. 2008), which, when overexpressed in vivo, was also protective against α-synuclein toxicity (Qiao et al. 2008; Cullen et al. 2009; Crabtree et al. 2014). Reduced expression of LAMP1 and cathepsin D was detected in nigral neurons with Lewy pathology (Chu et al. 2009). Importantly, heterozygous mutations in the lysosomal enzyme glucocerebrosidase (GBA) increase the risk of PD by fivefold (Sidransky et al. 2009; Nalls et al. 2013), which is at least partly a result of decreased degradation and increased toxicity of aggregated α-synuclein (Mazzulli et al. 2011; Sardi et al. 2011; Schöndorf et al. 2014). Pharmacological chaperones that promote GBA stability and trafficking to the lysosome (Steet et al. 2006; Khanna et al. 2010) were effective in improving lysosomal function and in reducing α-synuclein levels in cellular models (McNeill et al. 2014) and in Tg mice overexpressing human α-synuclein, where they improved motor function (Richter et al. 2014). It has also been reported that heterozygous GBA mutations in iPSC-derived neurons (Fernandez et al. 2016), as well as autophagic or lysosomal inhibition in cell lines (Alvares-Elviri et al. 2011), enhances the extracellular release of α-synuclein, and that GBA deficiency in Tg animals promotes the cell-to-cell transfer of α-synuclein aggregates (Bae et al. 2014). Collectively, these data suggest that α-synuclein is trafficked to lysosomes for degradation and that inhibition of this process can exacerbate intra- and transneuronal α-synuclein toxicity because of raised protein levels and secretion.

Since a fraction of α-synuclein in Lewy pathology is ubiquitinated (Tofaris et al. 2003; Anderson et al. 2006), ubiquitin ligases may be relevant to its trafficking and turnover. Nedd4 (neuronally expressed developmentally downregulated gene 4) serves a critical function in the endosomal–lysosomal pathway, promoting the degradation of membrane-associated proteins (Rotin and Kumar 2009). Nedd4 is downregulated during development and upregulated in response to oxidative stress (Hoshikawa et al. 2003), traumatic head injury (Sang et al. 2006), and neurodegeneration (Kwak et al. 2012), indicating that its expression in neurons is tightly regulated. Nedd4 and its yeast ortholog Rsp5 ubiquitinate α-synuclein in vitro by recognizing its proline-rich C-terminus (Tofaris et al. 2011). This leads to the conjugation of uniform K63-linked ubiquitin chains on specific lysine residues of α-synuclein (Lys-21 and Lys-96). Interestingly, Lys-96 of α-synuclein has been identified as its primary ubiquitination site in rat brain by an antibody-based proteomic analysis (Na et al. 2012). In addition, K63-linked conjugates are present in Lewy bodies, and their abundance inversely correlates with the de-ubiquitinase Usp8, which opposes the lysosomal degradation of α-synuclein in mammalian cells by de-ubiquitination and modifies its toxicity in the Drosophila model (Alexopoulou et al. 2016). In mammalian cells, Nedd4 overexpression promoted the degradation of α-synuclein by the ESCRT-mediated endosomal-lysosomal route and, in yeast, Rsp5 protected against α-synuclein toxicity (Tofaris et al. 2011). More recently, it was shown that Nedd4 overexpression protects against α-synuclein accumulation in Drosophila and rat models of α-synucleinopathy, whereas endogenous Nedd4 in Drosophila is especially critical for the dopaminergic neuronal response to α-synuclein toxicity (Davies et al. 2014). Nedd4 is upregulated in a subpopulation of pigmented neurons containing Lewy pathology (Tofaris et al. 2011), and Nedd4 mRNA levels are increased in brain regions with Lewy pathology (Dumitriu et al. 2012). In agreement with these observations, a chemical genetic screen has identified a small molecule that binds to and activates Nedd4 as a neuroprotective agent in yeast and iPSC-derived cortical neuronal models of α-synuclein toxicity (Chung et al. 2013). That the ubiquitin ligase activity of Nedd4 and Rsp5 is necessary for neuroprotection, at least partly by a direct effect on α-synuclein (Tofaris et al. 2011; Davies et al. 2014), has been confirmed and extended by recent studies, which have shown that Nedd4 also promotes the endosomal processing of internalized α-synuclein (Sugeno et al. 2014) and have identified Rsp5 mutants in yeast that enhance α-synuclein ubiquitination and clearance (Wijayanti et al. 2014). Additionally, α-synuclein-independent protective effects are also possible. For example, Nedd4 promotes the ubiquitination and recycling of AMPA receptors at the synapse (Hou et al. 2011). AMPA receptors are also regulated, downstream from Nedd4, by the retromer (Munsie et al. 2014), which has been implicated in familial PD (Vilariño-Güell et al. 2011). In addition, Nedd4 has been shown to ubiquitinate cytosolic misfolded proteins following heat stress (Fang et al. 2014) and misfolded A53T mutant α-synuclein (Davies et al. 2014), suggesting that it may also promote the clearance of intracellular aggregates. Therefore, enzymes that function in endosomal trafficking directly regulate α-synuclein levels by ubiquitination and/or modify its toxicity in different model systems.

CONCLUDING REMARKS

In this review, we have highlighted recent advances suggesting that aggregated α-synuclein has prion-like properties. It can spread from cell to cell by uptake and release, templating further aggregation that eventually leads to damage and cell death. Recent studies have shown that both endogenous and internalized α-synuclein is processed via the endosomal route, suggesting that enzymes that regulate this trafficking step could serve as a nodal point in α-synuclein recycling; although such clearance pathways act primarily as disposal mechanisms via the lysosome, they could potentially be “hijacked” by aberrant conformers to promote pathogenicity under certain conditions. For example, ubiquitin-mediated endocytosis of α-synuclein-receptor complexes may regulate the translocation of misfolded α-synuclein assemblies in endosomes. These species could in turn interact with endogenously trafficked protein in endosomal compartments (early endosomes or multivesicular bodies), triggering template amplification (as shown in the proposed model in Fig. 1). Interestingly, oligomerization of α-synuclein has been shown to occur within vesicles (Jang et al. 2010). The low pH of vesicular compartments may reduce the negative charge of the C-terminus of α-synuclein, resulting in an increase in its hydrophobicity, which is more permissive for aggregation (Uversky et al. 2001; McClendon et al. 2009; Buell et al. 2014). Misfolded species could be released intracellularly because of membrane rupture (Freeman et al. 2013), acting as seeds for the aggregation of cytosolic α-synuclein or extracellularly via exocytosis and possibly exosome release. This would be reminiscent of Aβ and prions, where endosomal processing appears to be critical for the generation of the toxic species, as well as for physiological trafficking (Bohm et al. 2015; Goold et al. 2015). This also suggests that there may be a therapeutic window within which activation of such pathways is beneficial. Understanding these molecular events will clarify the therapeutic potential of enzymes that regulate protein trafficking and degradation in synucleinopathies.