Synuclein is a relatively small protein characterized by the presence of five imperfect repeats of a KT(A)KE(Q)G(Q)V motif in the amino-terminal half, hydrophobic amino acids in the midportion, and an acidic C-terminal region (Fig. 4.2). The repeat region has been reported to be important for the association of synucleins with lipid membranes [17]. Three homologous proteins with the repeat sequence have been cloned and named α-, β-, and γ-synucleins; they are encoded by three independent genes [18]. β-Synuclein has 78 % homology and γ-synuclein has 58 % homology to α-synuclein in terms of DNA sequences. α-Synuclein and β-synuclein are colocalized in synaptic terminals in many regions of the brain, whereas γ-synuclein is expressed in particular kinds of neurons, such as dorsal root ganglion or non-neuronal tissues, and is overexpressed in breast cancer cells.

Electron microscopy combined with immunocytochemistry showed that α-synuclein is localized in presynaptic terminals, especially near synaptic vesicles [10]. The physiological roles of the protein remain to be clarified, but several functions have been suggested. It may be involved in (1) memory, learning, and plasticity of synapses, since expression levels of the homologue synelfin are increased only during a limited period of song learning in zebra finch [19], and (2) transport processes in vesicles, since it is localized in membranes and inhibits phospholipase D2, a regulator of the cytoskeleton and endocytosis [20]. A chaperon-like role of synuclein has also been proposed [21]. Based on analyses of knockout mice lacking both α- and β-synuclein, α-synuclein may contribute to the long-term regulation and/or maintenance of presynaptic function, as well as protection of nerve terminals against injury, and may exhibit this activity in conjunction with CSPα and SNARE proteins on the presynaptic membrane interface [22]. Since only a minor use-dependent alteration of dopamine release was observed in α-synuclein KO mice, α-synuclein is not essential for survival or for neurotransmitter release. These studies led to the idea that, consistent with its localization, α-synuclein may regulate synaptic plasticity (reviewed in [18]). No synuclein homologue has been identified in worm or fly.

To investigate the molecular mechanisms of Lewy body pathogenesis, it is essential to know what are the differences between normal α-synuclein and the abnormal form in Lewy bodies. Biochemical analyses of α-synuclein from DLB and control brains demonstrated that normal α-synuclein is mostly recovered in Tris and Triton-X-soluble fractions, whereas the abnormal form is detected in sarkosyl-insoluble fraction (which is soluble in 8 M urea, 6 M guanidine hydrochloride, and SDS). Mass spectrometric analysis of α-synuclein revealed that Ser129 is abnormally phosphorylated in the sarkosyl-insoluble form [23]. Phosphorylation site-specific antibodies indicated that more than 90 % of insoluble α-synuclein is phosphorylated at Ser129, while only about 5 % is phosphorylated in normal synuclein [23]. Casein kinases 1, 2 (CK1, CK2) and G protein-coupled receptor kinases 2, 5 (GRK2, GRK5) are candidate kinases for the in vitro phosphorylation, and major kinase activities in rat brain extracts were reported to be due to CK2 [24]. The level of phosphorylation seems to be higher in the striatum than in the hippocampus in mouse brain, and it was decreased up to 57 % by cold-water-induced stress [25]. The physiological role of Ser129 phosphorylation in the brain remains unclear, but it may be a degradation signal for the ubiquitin proteasome system or it may affect some interaction of other proteins.

Other posttranslational modifications such as nitration [26] have also been reported, but their biological significance is unknown. Ubiquitination is another important modification in the abnormal phosphorylated form of α-synuclein, as confirmed by both immunochemical and mass spectrometric analyses. Because the extent of ubiquitination is less than that of phosphorylation, ubiquitination may be a later event. Ubiquitination sites of α-synuclein in Lewy bodies in PD and DLB and also glial cytoplasmic inclusions in MSA have been determined by MS analysis to be Lys12, 21, and 23, with Lys 48 multi-ubiquitination [27]. These sites are different from those found in soluble α-synuclein, suggesting that aggregated proteins may be recognized as abnormal by the ubiquitin proteasome system of the cells [27, 28].

A much more prominent feature of abnormal α-synuclein in Lewy bodies is its conformation. Ultrastructurally, it has been shown that α-synuclein in Lewy bodies and Lewy neurites consists of filamentous or fibrous forms with a width of 10 nm, which are completely different from the normal unstructured form. Furthermore, the fibrils are thioflavin S-positive, a characteristic feature of amyloid fibrils, as well as Aβ and tau fibrils [29]. As discussed later, recombinant α-synuclein forms fibrils biochemically and morphologically similar to those in diseased brains in the absence of posttranslational modification. Furthermore, X-ray diffraction analysis demonstrated that α-synuclein is assembled into fibrils with cross-β structure [30].

There is continuing debate as to whether or not the phosphorylation is a primary and causative event for the accumulation of intracellular abnormal proteins, especially for tau protein in tauopathies. However, at least in the case of α-synuclein, the results of in vitro and in vivo studies strongly suggest that phosphorylation is not a primary event leading to accumulation. Instead, the conformational change of α-synuclein is thought to be the primary event in diseased brains.

α-Synuclein is natively unfolded, i.e., it is basically unstructured, with no significant secondary structures [31]. So, in considering the molecular mechanisms of PD and DLB pathogenesis and potential therapies, the key questions seem to be: “how does the unstructured protein form amyloid-like fibrils?”, “how do pathogenic mutations in the α-synuclein affect the conformation and induce fibril formation?”, “what are the mechanisms of toxicity of the fibrils?”, and “how can we inhibit fibril formation?”.

Purified recombinant α-synuclein can be easily converted to filamentous form by incubation at 37 °C with shaking at a high concentration. Conformational change from random coil to β-sheet structure can be observed by CD spectral analysis and thioflavin S fluorescence measurement. In contrast, no such conformational changes are observed with β-synuclein, whose amino acid sequence is 61 % homologous to that of α-synuclein, suggesting that the difference is critical for fibril formation. Actually, α-synuclein lacking a hydrophobic stretch of 12 amino acid residues (71VTGVTAVAQKTV82) does not form fibrils, and this sequence is essential for α-synuclein fibril formation [32]. As mentioned above, the fibrils have cross-β structure typical of amyloid fibrils [30]. MALDI-TOF mass analysis of proteinase K-resistant fragments of α-synuclein fibrils revealed that the core of α-synuclein fibrils is composed of residues 31–109 of α-synuclein [33].

To date, six missense mutations in the α-synuclein gene (A30P, E46K, H50Q, G51D, A53T, and A53E), as well as occurrence of gene duplication and triplication, have been identified in familial forms of PD and DLB [5, 34–41]. The A53T and E46K mutants form fibrils faster than wild-type α-synuclein (WT), but fibrillization of A30P is slow [42]. Instead, A30P mutant was suggested to form oligomeric protofibrils, and oligomer formation may the cause of PD [43]. However, we found that A30P mutant forms a fragile-type fibril, which resulted in accelerated nucleation-dependent fibrillization of α-synuclein compared with WT fibrils [44]. Later, autopsy of a patient with the A30P mutation revealed abundant α-synuclein pathology in the brain, strongly suggesting that the mutation induced PD via accumulation of pathological α-synuclein fibrils, not formation of soluble oligomers [45]. These results support the idea that αsyn fibril formation causes PD and DLB.

Although α-synuclein is natively unfolded, as mentioned above, NMR analysis of the monomer revealed interactions between NAC (residues 61–95) and the C-terminal region (residues 110–130), between residues 120–130 and 105–115, and between the N-terminal region and near residue 120 [46], indicating that the protein molecule is not totally unfolded. The natively unfolded state of α-synuclein may be stabilized by long-range interactions (e.g., shielding of the hydrophobic NAC domain by the acidic C-terminal region) that inhibit aggregation or fibrillization [46]. This could be the reason why C-terminally truncated forms have higher propensities to aggregate [47]. The A30P and A53T mutations were reported to influence the stability of the native state and to favor fibril formation [48]. The conformational changes of α-synuclein in going from the monomer state to amyloid-like fibrils can be detected more easily by using site-specific α-synuclein antibodies [49]. Antibodies against the N-terminal region or NAC region of α-synuclein do not react very well with the monomer, but strongly stain the fibrils in dot blot analysis, while antibodies against the C-terminal region stain both monomer and fibrils well [49].

Fibrillization or conformational change of α-synuclein is thought to be the central event in the pathogenesis of PD and DLB. Assembly of protein into amyloid fibrils is usually a nucleation-dependent process that consists of nucleation followed by elongation, and α-synuclein fibrillization was also confirmed to be nucleation dependent (Fig. 4.3) [50]. Addition of seeds (a small amount of preformed fibrils) to the monomer promotes fibrillization by skipping the nucleation process. Although the precise mechanism remains to be characterized, the conversion of monomer to fibrils seems to be basically a prion-like conversion. Namely, the conformation is changed to the abnormal form by interaction of the monomer with fibril seeds. Amyloid-like fibril seeds act as a template for conversion, and multiplication of this conversion induces propagation of well-ordered homogeneous amyloid-like fibrils.

This seed-dependent conversion and fibril formation of α-synuclein have been established in culture cells by introducing fibril seeds with transfection reagents. Aggregation or fibrillization does not occur when α-synuclein is overexpressed in cultured cells by means of plasmid transfection, even if pathogenic mutants are induced, or when cells are exposed to various stresses. In contrast, abundant α-synuclein inclusions were formed inside cells transfected with α-synuclein plasmid when preformed fibril seeds were introduced into the cells with a transfection reagent (Fig. 4.4) [51]. EM observation of the inclusions revealed that they were composed of 10-nm α-synuclein fibrils similar to those found in DLB brains. Biochemical and immunocytochemical analyses demonstrated that the aggregated α-synuclein is abnormally phosphorylated at Ser129 and partially ubiquitinated and is recovered in sarkosyl-soluble and sarkosyl-insoluble fractions; it is indistinguishable from the aggregates in DLB brains [51]. Interestingly, cells with inclusions have been shown to degenerate gradually during 3–4 days in culture. Monitoring of proteasome activities in cells using GFP-CL1 (a degron that is an effective proteasome-degradation signal) and additional biochemical analyses indicated that proteasome dysfunction induced by abnormal α-synuclein may be one of the major causes of cell necrosis [51]. Similar seed-dependent aggregation has been observed with other intracellular pathological proteins, such as tau and TDP-43 [51, 52].

Many lines of transgenic mice overexpressing wild-type α-synuclein, mutant α-synuclein, or truncated forms have been produced in attempts to recapitulate the neuropathologies of PD and DLB. However, only a few mouse models develop phosphorylated α-synuclein pathology. The M83 line (expressing mutant A53T human α-synuclein) of Tg mice develops severe movement disorder with neuronal α-synuclein pathologies [53]. The relative levels of overexpression of α-synuclein, compared to endogenous protein, are about 3.3 (in hemizygous mice) and 4.6 (in homozygous mice) in the cortex and 19.3 (in hemizygous mice) and 28.2 (in homozygous mice) in the spinal cord. Homozygous mice exhibit signs of spontaneous neurological dysfunction beginning at ~8 months of age [53]. Furthermore, the mice develop age-dependent intracytoplasmic neuronal α-synuclein inclusions that recapitulate features of the human counterparts. Immunoelectron microscopy revealed that the α-synuclein inclusions contain 10–16 nm wide fibrils similar to those in human pathological inclusions [53]. Thus, overexpression of mutant human α-synuclein leads to the intracellular accumulation of filamentous α-synuclein that induces neurodegeneration in these mice.

Recently, α-synuclein pathology was found to be accelerated by injection of preformed fibril seeds or insoluble fractions from aged M83 Tg-mouse brains into brains of young Tg mice [54], as previously found in tau Tg mice [55]. The pathologies were also detected in the un-injected hemisphere of the brain, suggesting that intracellular pathological proteins may spread from the injected area to the other hemisphere [54]. However, it is difficult to rule out the possibility that α-synuclein fibrils accelerate production of the abnormal form by some other mechanism(s), because Tg mice are predisposed to develop the pathology. However, strikingly, prion-like propagation of α-synuclein pathology in the brain was clearly demonstrated in non-Tg, C57B6 wild-type mice by Luk et al. [56] and our group (Fig. 4.5) [57]. Luk et al. injected preformed mouse α-synuclein into the striatum of wild-type mouse brain and demonstrated the development of phospho-α-synuclein pathology at 3 months after injection, together with neuronal loss and motor dysfunctions [56]. We also showed that inoculation of preformed human and mouse α-synuclein fibrils into substantia nigra of wild-type mouse brain induced phosphorylated and ubiquitinated α-synuclein pathology in various brain regions and that mouse α-synuclein fibrils induced the pathology more efficiently than human α-synuclein fibrils, suggesting there may be a species barrier, as has been found in the case of infection of prion proteins [57]. Importantly, our study clearly demonstrated that mouse endogenous α-synuclein is converted to an abnormal form that is indistinguishable from the form in DLB brain by inoculation of human α-synuclein fibrils, though the injected human α-synuclein fibrils are degraded within a week after injection [57]. We also showed that sarkosyl-insoluble pellets from DLB brains, in which abnormal α-synuclein fibrils are enriched, induced similar pathologies of phospho-α-synuclein [57], strongly suggesting that pathological α-synuclein protein spreads from cell to cell by prion-like propagation in DLB brains. No such pathology was detected in mice injected with soluble α-synuclein monomer or in α-synuclein KO (deficient) mice injected with α-synuclein fibrils. Recently, Watts reported that brain homogenates from MSA brains also accelerated and induced α-synuclein pathology and motor dysfunction in the heterozygous M83 Tg mice; since α-synuclein fibrils fulfilled all the criteria for classification as prions, the name “α-synuclein prions” was proposed [58].