https://orcid.org/0009-0008-3354-5417
Abstract
α-Synuclein (α-Syn) is a protein related to synucleinopathies with high expression in the central nervous system and erythrocytes which are a major source of peripheral α-Syn. Recent reports have suggested the presence of α-Syn within extracellular vesicles (EVs) derived from erythrocytes, potentially contributing to the pathogenesis of synucleinopathies. While Lewy bodies, intracellular inclusions containing aggregated α-Syn, are prominently observed within the brain, their occurrence in peripheral neurons implies the dissemination of synucleinopathy pathology throughout the body via the propagation of α-Syn. In this study, we found erythrocytes and circulating EVs obtained from plasma contained α-Syn, which was separated into four major forms using high-resolution clear native-PAGE and isoelectric focusing. Notably, erythrocyte α-Syn was classified into full-length and C-terminal truncated forms, with truncation observed between Y133 and Q134 as determined by LC–MS/MS analysis. Our finding revealed that C-terminally truncated α-Syn, which was previously reported to exist solely within the brain, was also present in erythrocytes and circulating EVs obtained from plasma.
α-Synuclein (α-Syn) is a protein related to synucleinopathies, including Parkinson’s disease (PD), dementia with Lewy bodies and multiple system atrophy (1). Comprising 140 amino acids, human α-syn is predominantly located in the central nervous system and is expressed systemically. Post-translational modification (PTM) of α-syn may lead to misfolding, and then the misfolded α-syn may behave as a core of aggregation to enhance neurodegeneration (2–6). In fact, α-syn aggregates are found with various PTMs, such as phosphorylation and C-terminal truncation in pathological inclusion extracts (7–9). Interestingly, similar aggregates have also been observed in individuals with Alzheimer’s disease (10–12) as well as those without neurodegenerative diseases (7). In pathological inclusion extracts, 15–20% of α-syn is reported to be in a truncated form (12), and in vitro experiments have suggested that C-terminal truncated α-syn has a higher propensity for fibrillization compared to full-length α-syn or mutant α-syn variants associated with familial PD (13–15). It has been suggested that truncated α-syn, which is believed to be formed due to age-related proteostasis impairment (16), may also contribute to the progression of Lewy body pathology (17–22). The molecular forms of α-syn in erythrocytes under physiological conditions remain unclear. Methodological differences in purification have led to α-syn being identified as either a monomer (23) or tetramer (24). In addition, increased levels of dimeric α-syn in erythrocyte membrane fractions have been noted (25), and patients with PD exhibit an elevated oligomeric/total α-syn ratio (26).
Predominantly localized within the presynaptic region of central nervous system neurons (27, 28), α-syn has also been identified in extracellular vesicles (EVs) secreted from neurons, microglia and astrocytes, serving as key agents in α-syn propagation (29, 30). Recent investigations have revealed that α-syn can be transported from the brain to the blood within EVs (31, 32), and it has also been detected in EVs isolated from the cerebrospinal fluid and serum (33, 34). Remarkably, α-syn is highly expressed in erythrocytes as well as the central nervous system (23). In the peripheral blood, over 90% of α-syn is present in erythrocytes, which are a major source of peripheral α-syn (35, 36). Intriguingly, CD235a-positive human erythrocyte-derived EVs have demonstrated bidirectional transport across the blood–brain barrier (37), and erythrocyte-derived EVs from patients with PD exhibited heightened protein oligomerization in the brains of PD model mice compared to those from healthy subjects (37). Notably, PD lesions exhibit a broader distribution beyond the confines of the central nervous system; α-syn inclusions are observed in their initial stages within the peripheral nervous system (38). α-Syn inclusions have been identified not only within the substantia nigra but also within the sympathetic ganglia, adrenal cortex and the gastrointestinal tract wall (39–41).
In this study, we elucidated the presence of the C-terminally truncated form of α-syn as a prominent PTM of erythrocyte α-syn. Furthermore, this truncated form was also observed to be released via EVs into plasma.
Materials and Methods
Blood specimens
This study was conducted according to the Declaration of Helsinki and applicable national laws and regulations and was approved by the Ethics Committee of the Faculty of Science, Toho University (27-1, 2019-5). Written informed consent was obtained from all the participants. In this study, healthy volunteers were recruited from Toho University. The platelet-poor plasma fraction was isolated by centrifugation, following a previously described method (42) with minor modifications. Whole blood was collected in EDTA-2Na tubes (Venoject II, Terumo), and the blood components were separated within 30 min by centrifugation at 1,000×g for 20 min at room temperature to prevent degradation and loss of EVs. The resulting supernatant was collected as the plasma fraction, leaving ~5 mm layer of plasma above the buffy coat untouched. The remaining plasma and buffy coat were corrected, washed with erythrocyte lysing buffer multiple times (43) and stored at −80°C for other research projects. Subsequently, the plasma fraction underwent additional centrifugation at 1,600×g for 15 min at room temperature to pellet larger cell debris and remove remaining platelets, aiming to isolate the platelet-poor plasma fraction for further isolation of EVs. Erythrocytes were then washed three times with phosphate-buffered saline (PBS). The cytosolic fraction of erythrocytes was isolated as previously described (44) and stored at −80°C until analysis.
Circulating EVs in plasma
Circulating EVs in plasma were isolated by ultracentrifugation, as previously described (45–47) with minor modifications. EV subtypes were classified with medium and large EVs (m/lEVs: > 200 nm) and small EVs (sEVs: < 100 or < 200 nm) according to the size of EVs, following minimal information for studies of EVs 2018 (48). The platelet-poor plasma fraction was centrifuged at 20,000×g for 20 min at 4°C. The pellet after 20,000×g centrifugation was collected as a m/lEVs fraction and washed three times with PBS at 20, 000×g for 20 min at 4°C. The sEVs fraction was collected by ultracentrifugation of the supernatant of the m/lEVs fraction at 187,000×g for 70 min at 4°C and washed twice with PBS at 187,000×g for 70 min at 4°C. The collected m/lEVs and sEVs samples were stored at −80°C until use. Circulating EVs in plasma were confirmed by western blotting.
Size-exclusion chromatography
Haemoglobin in the cytosolic erythrocyte fraction was eliminated by nickel affinity chromatography, as described previously (49), with minor modifications. Briefly, 100 mg of erythrocyte protein was applied into 5 ml of a Ni Sepharose 6 Fast Flow column (Cytiva) equilibrated with 25 mM potassium phosphate buffer (pH 7.4) containing 10 mM NaCl. Pass fractions containing α-syn were concentrated by ultrafiltration. The sample was then applied to a HiPrep 16/60 Sephacryl S-100HR column (Cytiva), equilibrated with PBS. In addition, 25 μg of recombinant human α-syn monomer expressed in Escherichia coli (ab218816; Abcam) was applied to the Sephacryl S-100HR column and fractionated.
High-resolution clear native (hrCN)-PAGE and two-dimensional hrCN/SDS-PAGE
Erythrocyte α-syn was separated by hrCN-PAGE as described previously (50), with minor modifications (51). For two-dimensional hrCN/SDS-PAGE analysis, 400 μg of the erythrocyte sample was applied onto a 13% hrCN-PAGE gel as the first dimension. After hrCN-PAGE, each lane was cut and incubated with denaturing buffer (2× SDS-PAGE sample buffer with 5 mM 2-mercaptoethanol) for 30 min at 37°C, followed by alkylation buffer (2× SDS-PAGE sample buffer with 54 mM N,N-dimethylacrylamide) for 15 min at 37°C. The gel strips were subjected to 15% SDS-PAGE in the second dimension.
Two-dimensional isoelectric focusing (IEF)/hrCN-PAGE
First-dimensional IEF was performed using a PROTEAN IEF CELL (Bio-Rad). For two-dimensional IEF/hrCN-PAGE analysis, 200 μg of erythrocyte protein was mixed with an IEF sample buffer (1632106, Bio-Rad) and applied on a 1D gel strip (pH 3–6, 1632003, Bio-Rad). After focusing, the gel strip was incubated with 2× hrCN-PAGE sample buffer (100 mM Bis–Tris [pH 7.0], 100 mM sodium chloride, 20% glycerol and 0.002% Ponceau S) for 30 min at room temperature and then fixed on a 13% hrCN-PAGE gel with 4% stacking gel solution. α-Syn was detected by Western blotting, and an isoelectric point (pI) was calculated.
Western blotting
hrCN-PAGE gels were incubated with denaturing buffer as described previously (51), and the sensitive western blotting method (52) was performed. α-syn was detected with the following antibodies: 1:10,000 dilution anti-α-syn antibody [MJFR1] (ab138501, aa 118–123, Abcam), 1:1,000 dilution anti-α-syn antibody [4D6] (ab1903, aa 124–134, Abcam), 1:5,000 dilution anti-α-syn antibody clone Syn211 (S5566, aa 121–125, Sigma-Aldrich), 1:5,000 dilution anti-α-Syn 1–10 (TIP-SN-P01, aa 1–10, COSMO BIO), 1:5,000 dilution anti-α-Syn 75–91 (TIP-SN-08, aa 75–91, COSMO BIO) and 1:5,000 dilution anti-α-Syn 131–140 (TIP-SN-09, aa 131–140, COSMO BIO). The epitopes of the 4D6 antibody were mapped according to a previous study (53), and the other antibodies were disclosed by their respective suppliers. Glycophorin A was detected using 1:2,500 dilution anti-glycophorin A antibody (66778-1-lg, Proteintech). CD81 was detected using 1:500 dilution anti-CD81 antibody (sc-7637, Santa Cruz Biotechnology). All antibodies were diluted in tris buffered saline with Tween-20 (TBS-T) and an EzWest Lumi plus (ATTO) was used for horseradish peroxidase detection. Signal detection was performed using ImageQuant LAS 4010 (Cytiva) or Amersham Imager 680QC (Cytiva).
LC–MS/MS analysis
Erythrocyte α-syn was purified by immunoprecipitation using anti-α-syn antibody [MJFR1] (Abcam), as described previously (44). LC–MS/MS analysis for the identification of PTMs was performed as described previously with a modified protocol (44, 54, 55). The specified search parameters were as follows: database, SwissProt (2019_11, Swiss Institute of Bioinformatics), taxonomy (Homo sapiens), static modification (carbamidomethyl [C]) and variable modification (oxidation [M]).
Results
C-terminal truncation between Y133 and Q134 was a prominent PTM observed in erythrocyte α-syn
We first investigated the reactivity of anti-α-syn antibodies against erythrocyte α-syn by hrCN-PAGE/western blotting using the following anti-α-syn antibodies: α-Syn 1–10, α-Syn 75–91, MJFR1, Syn211, 4D6 and α-Syn 131–140 (Fig. 1). α-Syn species were identified as bands A, B, C and D using α-Syn 1–10, α-Syn 75–91, MJFR1 and Syn211 antibodies, whereas bands A and B were not detected when 4D6 and α-Syn 131–140 antibodies were used (Fig. 1).
α-Syn species in the cytosolic fraction of erythrocytes. α-Syn species in erythrocytes were separated by hrCN-PAGE and visualized by western blotting using the following anti-α-syn antibodies: α-Syn 1–10, α-Syn 75–91, MJFR1, Syn211, 4D6 and α-Syn 131–140. α-Syn bands A–D are shown.
Subsequently, we compared the molecular forms of the recombinant human α-syn monomer (Abcam) and erythrocyte α-syn by size-exclusion chromatography. The fractions obtained from recombinant α-syn and haemoglobin-eliminated cytosolic erythrocyte proteins were individually subjected to SDS-PAGE (Fig. S1). Erythrocyte α-syn was fractionated within the volume range of 50–55 mL, which closely paralleled that of recombinant α-syn (Fig. S1). Then, fractions extracted from 50 to 55 mL of the haemoglobin-eliminated erythrocyte protein were subjected to hrCN-PAGE, resulting in the elution of all four erythrocyte α-syn species within the same fraction (Fig. 2A). Likewise, the erythrocyte protein, the recombinant α-syn monomer (Abcam) and the recombinant human α-syn N-terminal acetylated monomer (StressMarq) were subjected to hrCN-PAGE for a mobility comparison (Fig. 2B). Irrespective of N-terminal acetylation, the recombinant α-syn monomer exhibited the same mobility pattern as band C of erythrocyte α-syn.
Molecular forms of human α-syn. (A) Haemoglobin-eliminated erythrocyte protein was separated using a Sephacryl S-100HR. Each fraction was subjected to hrCN-PAGE and western blotting using the MJFR1 antibody. (B) Cytosolic protein of erythrocytes (lane 1), recombinant human α-syn monomer (lane 2) and recombinant human α-syn N-terminal acetylated monomer (lane 3) were subjected to hrCN-PAGE, and α-syn was visualized by Western blotting using the MJFR1 antibody.
To further analyse the bands A–D observed on the hrCN-PAGE, a second-dimensional SDS-PAGE was employed (Fig. 3A). In this analysis, bands A, B, C and D were detected as distinct spots labelled as a, b, c and d, respectively, and their molecular masses were determined to be 15, 15, 16 and 16 kDa, correspondingly. Furthermore, erythrocyte proteins underwent a two-dimensional IEF/hrCN-PAGE separation (Fig. 3B). The α-syn was separated into four spots, designated A, B, C and D, on two-dimensional hrCN-PAGE. Notably, these spots demonstrated consistent mobilities on the hrCN-PAGE separation, despite the inclusion of urea, 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate (CHAPS) and dithiothreitol in the IEF sample buffer. The pI values were measured in triplicate for a healthy subject, yielding pI values of 4.88 (0.006), 4.78 (0.012), 4.67 (0.025) and 4.59 (0.012) for spots A–D, respectively (Table I).
Two-dimensional PAGE of α-syn in erythrocyte protein. (A) Erythrocyte protein was analysed by 2D-hrCN/SDS-PAGE. α-Syn was probed with the MJFR1 antibody. α-Syn bands (A–D) on first-dimensional hrCN-PAGE and the spots on second-dimensional SDS-PAGE (a–d) are shown. (B) Erythrocyte protein was analysed by 2D-IEF/hrCN-PAGE. α-Syn was probed with the MJFR1 antibody. α-syn spots (A–D) on second-dimensional hrCN-PAGE are shown.
. Isoelectric point of erythrocyte α-synuclein
| Spot | 1 | 2 | 3 | Mean | SD |
|---|---|---|---|---|---|
| A | 4.88 | 4.87 | 4.87 | 4.87 | 0.006 |
| B | 4.80 | 4.78 | 4.78 | 4.78 | 0.012 |
| C | 4.70 | 4.65 | 4.66 | 4.67 | 0.025 |
| D | 4.60 | 4.58 | 4.58 | 4.58 | 0.012 |
| Spot | 1 | 2 | 3 | Mean | SD |
|---|---|---|---|---|---|
| A | 4.88 | 4.87 | 4.87 | 4.87 | 0.006 |
| B | 4.80 | 4.78 | 4.78 | 4.78 | 0.012 |
| C | 4.70 | 4.65 | 4.66 | 4.67 | 0.025 |
| D | 4.60 | 4.58 | 4.58 | 4.58 | 0.012 |
Isoelectric point values of erythrocyte α-synuclein were in the triplicate, and these of spots A–D in Fig. 3B were shown.
. Isoelectric point of erythrocyte α-synuclein
| Spot | 1 | 2 | 3 | Mean | SD |
|---|---|---|---|---|---|
| A | 4.88 | 4.87 | 4.87 | 4.87 | 0.006 |
| B | 4.80 | 4.78 | 4.78 | 4.78 | 0.012 |
| C | 4.70 | 4.65 | 4.66 | 4.67 | 0.025 |
| D | 4.60 | 4.58 | 4.58 | 4.58 | 0.012 |
| Spot | 1 | 2 | 3 | Mean | SD |
|---|---|---|---|---|---|
| A | 4.88 | 4.87 | 4.87 | 4.87 | 0.006 |
| B | 4.80 | 4.78 | 4.78 | 4.78 | 0.012 |
| C | 4.70 | 4.65 | 4.66 | 4.67 | 0.025 |
| D | 4.60 | 4.58 | 4.58 | 4.58 | 0.012 |
Isoelectric point values of erythrocyte α-synuclein were in the triplicate, and these of spots A–D in Fig. 3B were shown.
To determine the C-terminal truncation site using LC–MS/MS, α-syn was immunoprecipitated using the MJFR1 antibody. Subsequently, α-syn was separated into two distinct bands with SDS-PAGE, viz., full-length (16 kDa) and C-terminally truncated (15 kDa) forms (Fig. S2). Each of these bands was excised and digested with proteases, as described previously (44), for LC–MS/MS analysis. The sequence coverage of full-length α-syn was 72% (aa 11–97 and 126–140) by trypsin + chymotrypsin digestion and 87% by Glu-C digestion (aa 1–114 and 124–139). The C-terminally truncated α-syn sequence revealed aa 11–97 and 103–133 by trypsin + chymotrypsin digestion and aa 2–114 and 123–131 by Glu-C digestion (Fig. 4). All peptide fragment sequences and details from the Mascot search are shown in supplementary tables (Tables S1–S4).
Reading sequences of α-synuclein with LC–MS/MS analysis and digestion sites with trypsin + chymotrypsin and Glu-C. The sequence of full-length and C-terminally truncated α-synuclein obtained by LC–MS/MS is shown separately for trypsin + chymotrypsin and Glu-C. Read sequences are indicated by shading and the added marks indicate the theoretical digestion sites of each protease.
Circulating EVs in plasma contain both full-length and C-terminally truncated α-syn
The source of EVs containing α-syn in plasma within 1 mL of plasma was determined by SDS-PAGE/Western blotting using anti-CD81, anti-CD235a and anti-α-syn antibodies (Fig. 5A). The results showed that the anti-CD235a antibody, which is used as an erythrocyte EV marker that is highly expressed on the erythrocyte membrane, the anti-α-syn antibody reacted with a fraction pelleted at 20,000×g, and the anti-CD81 antibody, which is used as a typical exosome marker, reacted with the fractions pelleted at both 20,000×g and 187,000×g. These data demonstrate that erythrocyte α-syn was present in m/lEVs (e.g. microvesicles, MVs). In order to analyse the molecular forms of α-syn in m/lEVs, we performed a two-dimensional IEF/hrCN-PAGE (Fig. 5B). α-Syn contained in m/lEVs within 24 ml of plasma was separated into four spots, A, B, C and D, on two-dimensional hrCN-PAGE, which was the same pI values of erythrocyte α-syn.
Circulating EVs in plasma contain α-syn species. (A) Medium and large extracellular vesicles (m/lEVs) and small EV (s/EVs) fractions within 1 ml of plasma were subjected onto non-reduced SDS-PAGE. The m/lEVs marker protein CD235a, the sEVs marker protein CD81 and α-syn were detected by western blotting using specific antibodies. (B) Medium and large EVs (m/lEVs) fraction within 24 ml of plasma was analysed by 2D-IEF/hrCN-PAGE. α-Syn was probed with the MJFR1 antibody. α-Syn spots (A–D) on second-dimensional hrCN-PAGE are shown.
Discussion
The presence of C-terminally truncated α-syn has been reported in brain autopsies of individuals with synucleinopathies (7). While several studies have focused on erythrocyte α-syn as a biomarker for PD (25, 56–59), the molecular forms and PTMs have remained elusive. In this study, we demonstrated that α-syn in the cytosolic fraction of erythrocytes was separated into four bands on hrCN-PAGE. Notably, hrCN-PAGE is an improved method of blue native PAGE (50, 51). The micelles of the anionic and neutral detergents in the running buffer impose a negative charge shift on membrane proteins; consequently, negatively charged proteins can be separated based on their apparent molecular masses. On the other hand, soluble proteins in hrCN-PAGE are affected by molecular mass, conformation or electric charge. Gould et al. (60) have shown that α-syn in the human brain has conformationally diverse metastable conformers on clear native PAGE, which involves no detergents in the running buffer of hrCN-PAGE. They used anti-α-syn antibodies that recognized aa 2–12, 61–85 and 121–125 of α-syn. Interestingly, our findings, pertaining to human erythrocyte α-syn separation via hrCN-PAGE, are similar to those from brain α-syn separation by clear native PAGE. Araki et al. conducted a comparative analysis between human erythrocyte α-syn and human recombinant α-syn expressed in E. coli, revealing no significant disparities in molecular weight or secondary structure. Notably, the protein radius of gyration was influenced by buffer conditions, and the absence of a tetramer was observed (49). In our study, the employment of Sephacryl S-100HR separation revealed concurrence between the elution fraction of the four erythrocyte α-syn bands from hrCN-PAGE and the human recombinant α-syn monomer expressed in E. coli (Fig. 2A; Fig. S1). Interestingly, the recombinant human α-syn corresponded in mobility to band C of erythrocyte α-syn within hrCN-PAGE separation (Fig. 2B).
Our results showed that Y133-truncated α-syn (aa 1–133) is a prominent PTM of human erythrocytes. The antibodies recognizing aa 124–136 and aa 131–140 failed to detect the low-mobility bands A and B on hrCN-PAGE separation, anticipated as C-terminally truncated α-syn. In contrast, these bands were recognized by antibodies targeting aa 1–125 (Fig. 1). Employing two-dimensional hrCN/SDS-PAGE analysis, we observed the convergence of bands A–D from the first-dimensional hrCN-PAGE into spots a, b (15 kDa) and c, d (16 kDa) on the second-dimensional SDS-PAGE (Fig. 3A). This observation leads us to categorize bands A and B as truncated forms and bands C and D as full-length forms. Our LC–MS/MS analysis revealed the absence of the sequence of aa 134–140 (QDYEPA) in the C-terminally truncated form on SDS-PAGE (Fig. 4; Tables S2 and S4). Two-dimensional IEF/hrCN-PAGE analysis additionally demonstrated that the variation between spots A–D on hrCN-PAGE separation was attributed to differences in pI (Fig. 3B; Table I). Utilizing the Expasy Compute pI/MW tool (Swiss Institute of Bioinformatics), we calculated theoretical pI values for full-length and C-terminally truncated α-syn: 4.67 for aa 1–140 and 1–139, 4.72 for aa 1–138 and 1–137, 4.78 for aa 1–136 and 1–135 and 4.87 for aa 1–134, 1–133, 1–132 and 1–131, respectively (Table S5). Importantly, the estimated pI value of band C (4.67) aligned with the theoretical pI of full-length α-syn, and the estimated pI of band A (4.87) correlated with the theoretical pI values of aa 1–131, 1–132, 1–133 and 1–134 (Table I and Table S5). Furthermore, the estimated pI values for bands B and D, regarded as variants of bands A and C, were 4.78 and 4.58, respectively (Table I). These bands, B and D, could signify molecular forms experiencing PTMs that influence the charge of erythrocyte α-syn. Our prior work has indicated that lysine residues within erythrocyte α-syn serve as major targets for various PTMs (44). For instance, acetylation effects were explored utilizing recombinant mutants, where lysine residues were substituted with glutamine to simulate acetyl lysine (61). In addition to the truncation site, we estimated pI values by substituting K with Q residues using the Compute pI/MW tool (Table S6). This computation resulted in a pI shift from 4.67 to 4.59 for full-length α-syn and from 4.87 to 4.77 for Y133 truncated α-syn. These calculated values were closely paralleled the measured values via IEF/hrCN-PAGE (Table I). These findings collectively suggest that truncation at Y133 and lysine residues masking through PTMs can influence the pI of erythrocyte α-syn. Consequently, hrCN-PAGE revealed the isolation of four major bands. However, PTM-specific antibodies for lysine residues, such as anti-acetyl-lysine, anti-di-glycyl lysine, anti-ubiquitin and anti-SUMO1 did not recognize bands B and D (data not shown). Developing and validating new antibodies targeting specific α-syn PTM site would be a valuable avenue for further exploration. Therefore, we conclude that Y133 truncation constitutes a prominent PTM in human erythrocyte α-syn, observed ubiquitously even among healthy individuals.
Y133 truncation has also been detected in Lewy bodies, alongside α-syn species encompassing aa 1–115, 1–119, 1–122 and 1–135, as confirmed by MS (7). In an extensive review investigating truncated α-syn, a combination of epitope mapping, MS and truncation-specific antibodies was employed to examine C-truncated variants of α-syn within detergent-insoluble disease fractions. These analyses revealed truncation sites at residues N103, E110, L113, E114, D115, D119, N122, A124, Y125, Y133 and D135 (12). In the insoluble brain fraction, ~15–20% of α-syn was truncated (12), with aa1–119 and 1–122 truncations being predominant, detectable of 12–15 kDa bands on SDS-PAGE (12). In our study, the Y133-truncated α-syn is depicted as a 15 kDa band (Fig. 2S), comprising roughly 20% of the total α-syn bands with 15 kDa and 16 kDa, as analysed by ImageQuant TL Version 8.2 software. Moreover, truncated α-syn has been reported to be centrally located within inclusion bodies compared to full-length α-syn, suggesting its potential role in the early stages of inclusion body formation (62–64). Furthermore, several proteases have been identified for truncating or completely degrading both physiological and misfolded α-syn (12). Notably, matrix metalloproteases and a zinc-dependent endopeptidase family are implicated in neurodegenerative conditions. In particular, matrix metalloprotease-1, elevated in the brain tissue of patients with PD, catalysed Y133 truncation in recombinant human α-syn in vitro (65, 66). Moreover, the resultant Y133 truncation displayed a molecular mass of 15 kDa on SDS-PAGE and exhibited an enhanced aggregation propensity (65). Similarly, cathepsin D cleaved E123 of human α-syn, yielding 10- and 12-kDa proteins on SDS-PAGE (67). A calcium-dependent papain-like enzyme also performed truncations at E114 or N122, leading to facile α-syn fibrillation (20). Considering these findings, proteases present in erythrocytes might be responsible for C-terminal truncation of erythrocyte α-syn. However, the triggering factor behind this truncation remains undetermined and necessitates further investigation. The Y133-truncated α-syn we identified may share characteristics with the in vitro-aggregation-prone Y133-truncated α-syn, as reported previously (65). Although we have not verified its aggregation tendencies, C-terminal truncated α-syn is released into plasma via circulating EVs, in conjunction with full-length α-syn (Fig. 5B). Although the cleavage site of Y133-truncated α-syn differs from those reported for D119 and N122-truncated α-syn, which are reported to be more abundant in the brain (12), it is possible that further cleavage may occur after transit to the brain. Further investigations are required to understand the relationship between C-terminal truncated α-syn in erythrocytes and the pathogenesis of synucleinopathies.
Our observations showed that circulating m/lEVs in plasma contained erythrocyte-derived MVs, with no detection in CD81-positive sEVs derived within 1 ml of plasma (Fig. 5A). Lawrie et al. (68) employing dynamic light scattering, reported that unfiltered plasma-derived EVs range from 100 to 460 nm, which aligns with the absence of detection in CD81-positive sEVs (Fig. 5A). Regarding erythrocyte-derived EVs in plasma, it has been reported, based on analysis using freeze-fracture transmission electron microscopy, that the majority of erythrocyte-derived EVs belong to the size range of 120–200 nm (69). In another study employing atomic force microscopy, it was indicated that platelet-derived vesicles in plasma have an approximate size of 125 ± 21 nm (70). Recently, two separate studies have reported on the proportions of circulating MVs in the plasma of healthy volunteers. Through selective fluorochrome-labelled immunostaining, it was reported that platelet-derived MVs accounted for 88.0%, erythrocyte-derived MVs for 3.8%, endothelial-derived MVs for 0.5%, and unidentified MVs for 7.6% (71). In contrast, employing high-sensitivity flow cytometry, another study reported the proportions of various MV types as follows: platelet-derived MVs at 26%, erythrocyte-derived MVs at 22%, endothelial-derived MVs at 19%, granulocytes-derived MVs at 18%, leukocytes-derived MVs at 12% and monocytes-derived MVs at 4% (72). On the other hand, the amount of α-syn in the blood has been reported at ~26,200 ng/ml in erythrocytes, 50 ng/ml in platelets, 13.2 ng/ml in peripheral blood mononuclear cells and 25.4 ng/ml in plasma as measured by ELISA (35). These findings imply that if the quantity of α-syn released into MVs correlates with the intracellular content, MVs originating from erythrocytes might potentially contribute a higher α-syn content into the plasma compared to MVs derived from other sources. Further analysis is required to elucidate the molecular conformation and PTMs of α-syn present in alternative cells.
It has been reported using ELISA that the levels of haemoglobin–α-syn complex increased in erythrocyte and the cytoplasm of postmortem striatum and substantia nigra in ageing human (73). Interestingly, the levels of the haemoglobin–α-syn complex in the mitochondria of striatum and substantia nigra decreased with age. In contrast, the levels of the haemoglobin–α-syn complex in the cerebellum, whether in the cytoplasm or mitochondria, remained unchanged with age. The expression of haemoglobin is not restricted to erythrocytes but it is also present in neurons (74). In postmortem brain samples, it is unclear whether the haemoglobin in the haemoglobin–α-syn complex originates from neurons or erythrocytes. From the analysis of proteins within erythrocyte-derived MVs induced by calcium ionophores using nanoLC-MS/MS, it was determined that α-syn and haemoglobin were present among the constituents (72). This outcome aligns with prior studies that have reported the presence of α-syn in erythrocyte-derived MVs (31, 37, 72). However, it remains uncertain whether haemoglobin and α-syn form a complex within erythrocyte-derived MVs.
In summary, our investigation has unveiled the subdivision of erythrocyte α-syn into four major molecular forms through hrCN-PAGE and IEF, subsequently categorized as full-length and C-terminal truncated species via two-dimensional hrCN/SDS-PAGE. Furthermore, we have demonstrated the release of both full-length and C-terminal truncated α-syn via m/lEVs into plasma. Our results provide novel insight into the PTMs affecting erythrocyte α-syn and suggest that C-terminal truncated erythrocyte α-syn may be a new therapeutic target for neurodegenerative disease treatments.
Acknowledgements
This work was supported in part by JSPS KAKENHI Grant Number JP21K06861 (A.O.-M, R.S., T.T.), a project research grant from Toho University an initiative to realize diversity in the research environment (A.O-M), and the Medical Institute of Bioregulation Kyushu University Cooperative Research Project Program (T.T.).
Author Contributions
R.A.: Data curation, formal analysis, visualization and writing original draft; R.A., R.O., S.Y., H.N. and N.H.: investigation; R.S., T.T. and A.O.-M.: resources and funding acquisition; T.T.: data curation, methodology and writing—review and editing; A.O.-M.: conceptualization, data curation, methodology, project administration, supervision, visualization and writing—original draft, review and editing.
Supplementary Data
Supplementary Data are available at JB Online.
Conflict of Interest
None declared.
References
Author notes
Present address: Neurology Clinic Tsudanuma and Dowakai Chiba Hospital, Funabashi, Chiba 274-0825, Japan
