PHOSPHORYLATED α-SYNUCLEIN AT Ser129 IS TARGETED TO THE PROTEASOME PATHWAY IN A UBIQUITIN-INDEPENDENT MANNER

α-Synuclein (a-Syn) is a major component of fibrillar aggregates in Lewy bodies (LBs), a characteristic hallmark of Parkinson disease. Almost 90% of a-Syn deposited in LBs is phosphorylated at Ser-129. However, the role of Ser-129-phosphorylated a-Syn in the biogenesis of LBs remains unclear. Here, we investigated the metabolism of Ser-129-phosphorylated a-Syn. In SH-SY5Y cells, inhibition of protein phosphatase 2A/1 by okadaic acid, and inhibition of the proteasome pathway by MG132 or lactacystin accumulated Ser-129-phosphorylated a-Syn. However, these inhibitions did not alter the amounts of total a-Syn within the observation time. Inhibition of the autophagy-lysosome pathway by 3-methyladenine or chloroquine accumulated Ser-129-phosphorylated a-Syn in parallel to total a-Syn during longer incubations. Experiments using cycloheximide showed that Ser-129-phosphorylated a-Syn diminished rapidly (t(½) = 54.9 ± 6.4 min), in contrast to the stably expressed total a-Syn. The short half-life of Ser-129-phosphorylated a-Syn was blocked by MG132 to a greater extent than okadaic acid. In rat primary cortical neurons, either MG132, lactacystin, or okadaic acid accumulated Ser-129-phosphorylated a-Syn. Additionally, we did not find that phosphorylated a-Syn was ubiquitinated in the presence of proteasome inhibitors. These data show that Ser-129-phosphorylated a-Syn is targeted to the proteasome pathway in a ubiquitin-independent manner, in addition to undergoing dephosphorylation. The proteasome pathway may play a role in the biogenesis of Ser-129-phosphorylated a-Syn-rich LBs.

Sporadic Parkinson's disease (sPD) is characterized pathologically by a loss of dopaminergic neurons in the substantia nigra pars compacta and the presence of intracytoplasmic inclusions called Lewy bodies (LBs) and Lewy neurites (LNs) in surviving neurons. α-Synuclein (a-Syn) is a major component of fibrillar aggregates in LBs and LNs. Accumulating lines of evidence have shown that prefibrillar intermediates of a-Syn, such as soluble oligomers or protofibrils, play a toxic role in degeneration of dopaminergic neurons, and mature fibrils of a-Syn contribute towards this toxicity to a lesser extent (1)(2)(3)(4). Therefore, the process of a-Syn aggregation eventually forming LBs is proposed to play a causative role in neuronal degeneration of PD (5,6). Immunohistochemical and biochemical studies have revealed that approximately 90% of a-Syn deposited in LBs is phosphorylated at serine 129 (Ser129) (7,8). In contrast, the portion of phosphorylated a-Syn in normal brains is known to be only about 4% (7) or less than the limits of quantification of the assays used (8). This discrepancy implicates a pathogenic role of Ser129-phosphorylated a-Syn in the biogenesis of LBs (7,9). One possibility is that the Ser129-phosphorylation promotes the aggregation-prone property of a-Syn. To elucidate this issue, several in vitro studies have been performed. However, the accelerating effect of phosphorylation on fibril formation of a-Syn is controversial at present (7,10). Another possibility is that the impairment of the system to degrade Ser129-phosphorylated a-Syn causes its accumulation. However, the process by which Ser129-phosphorylated a-Syn is degraded or recycled remains unknown. This study focused on the metabolic fate of Ser129-phosphorylated a-Syn in cells. We report here that Ser129-phosphorylated a-Syn undergoes dephosphorylation and degradation by the proteasome pathway.
In addition, Ser129-phosphorylated a-Syn is targeted to the proteasome pathway in a ubiquitin-independent manner.

Experimental Procedures
Plasmid cDNA construction and reagents-Wild-type human a-Syn cDNA was described previously (11). Ser129Ala, Ser129Glu, Ser129Asp mutant a-Syn cDNAs were made by the two step PCR mutagenesis method. Ser9Ala/Ser42Ala/Ser87Ala mutant (it abolished Ser residues other than Ser129) and Lys12Arg/Lys21Arg/Lys23Arg mutant a-Syn cDNAs (it abolished Lys residues for ubiquitination) (8) were generated by applying the two-step PCR mutagenesis method. Human ubiquitin cDNAs with or without a FLAG-tag at the N-terminus were generated by PCR (clone ID #3879581; Open Biosystems, Huntsville, AL, USA), and they were subcloned into the pcDNA3.1 vector (Invitrogen, Carlsbad, CA, USA). Nucleotide sequences of all constructs were confirmed by direct sequencing. All reagents were purchased from Sigma (St. Louis, MO, USA) unless otherwise stated. Cell culture and transfection-Human dopaminergic neuroblastoma SH-SY5Y cells (ECACC #94030304) were maintained in a mixture of F-12 and Eagle's minimum essential medium supplemented with 15% fetal bovine serum (Invitrogen), 1  non-essential amino acids, and 2 mM L-glutamine (Invitrogen) at 37°C in 5% CO 2 . The SH-SY5Y cell line stably expressing wild-type a-Syn (wt-aS/SH) was selected against with 1 mg/ml G418 (Invitrogen). For transient transfection, 5  10 6 cells were transfected with 6 μg of cDNA using Nucleofector (Amaxa Cell Line Nucleofector kit V; Lonza Cologne AG, Koln, Germany). The cells were harvested at 48 h post-transfection. Primary neuronal cultures-Primary cortical neuron cultures were prepared from Crl:CD(SD) rats as previously described (11). Briefly, neurons were isolated from the neocortex of embryonic day 18 rats and dissociated cells were plated at a density of 1  10 6 cells on poly-D-lysine-coated 6-well plates (Becton Dickinson, Bedford, MA, USA). Neurons were maintained in serum-free medium, which was composed of neurobasal medium supplemented with B27 and GlutaMAX (Invitrogen). At intervals of two days, half of the plating medium was renewed. At 21 days of culture, neurons were harvested for experiments (12). Chemical treatments-For inhibition of the proteasome, at 16 h after plating wt-aS/SH cells onto 6-well plates, we checked the cells to be approximately 80 % confluent, and then the cells were further incubated in fresh medium containing either 10 nM okadaic acid, 10 μM MG132 or 10 μM lactacystin for 4 h. As a vehicle control, cells were treated with the same concentration of DMSO, which was used for dissolving okadaic acid, MG132 and lactacystin, or phosphate buffered saline (PBS: 1.06 mM KH 2 PO 4, 2.97 mM Na 2 HPO 4 ·7H 2 O, 150 mM NaCl). In rat primary neuronal cultures, neurons were cultured for 21 days and then incubated in fresh medium containing either 10 nM okadaic acid, 10 μM MG132 or 10 μM lactacystin for 4 h. For inhibition of the autophagy-lysosome, at 16 h post-plating wt-aS/SH cells onto 6-well plates, we confirmed the cells to be around 50% confluent. The cells were incubated in fresh medium containing 10 mM 3-metyladenine (3-MA) or 100 μM chloroquine for up to 32 h.
To assess protein half-lives in the cells, we performed experiments using the de novo protein synthesis inhibitor, cycloheximide. At 16 h post-plating wt-aS/SH cells onto 6-well plates, we confirmed the cells to be approximately 80% confluent. The cells were incubated in fresh medium containing 100 μM cycloheximide for the indicated times. To test the effect of inhibition of the proteasome pathway or dephosphorylation on the half-lives of target proteins, we treated the cells with cycloheximide plus either MG132 or okadaic acid. The 80% confluent wt-aS/SH cells were pre-incubated in fresh medium containing either DMSO, 10 μM MG132 or 10 nM okadaic acid for 6 h. After pre-incubation, cycloheximide was added to a final concentration of 100 μM into medium. The cells were further incubated for the indicated times. Protein extract preparation-For preparation of cell lysates, SH-SY5Y cells were suspended in buffer A [20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% NP-40, 10% glycerol, 1  protease inhibitor cocktail (Roche Diagnostic, Mannheim, Germany), 1 mM EDTA, 1  phosSTOP (Roche Diagnostic)] and kept on ice for 30 min. After centrifugation at 12,000  g for 30 min, the resultant supernatant was collected and stored at -80°C until required. In primary neuronal cultures, the cells were suspended in buffer A containing 1 μM okadaic acid, and then they were disrupted by passing through a 27G-needle 10 times.
In the experiments for ubiquitinated proteins, the pellet from above (post-12,000  g centrifugation step) was resuspended in the same aliquots of buffer A containing 8 M urea and was disrupted by brief sonication. After centrifugation at 12,000  g for 30 min, the resultant supernatant was collected as the pellet fraction. The protein concentration was measured by the BCA assay (Thermo Scientific, Rockford, IL, USA). Immunoprecipitation-The cells were suspended in ice-cold lysis buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% triton X-100, 0.1% SDS, 0.5% deoxycholic acid, 10% glycerol, 1  protease inhibitor cocktail, 1 mM EDTA, 20 mM NaF, 1 mM Na 3 VO 4, 1 μM okadaic acid), and kept on ice for 30 min. After centrifugation at 12,000  g for 30 min, the resultant supernatant was collected. The supernatants were incubated with primary antibodies overnight at 4°C, and then incubated with Protein-G agarose beads for 2 h. Beads were washed three times with ice-cold lysis buffer, and immunoprecipitates were dissolved from the beads by heating in Laemmli's sample buffer. Equivalent amounts of samples were analyzed by immunoblotting. Immunoblotting-For SDS-PAGE, protein samples were denatured at 95°C for 5 min in Laemmli's sample buffer containing 2.5% 2-mercaptoethanol. Samples were applied to a 12.5% SDS-polyacrylamide gel, electrophoresed and then transferred to a PVDF membrane (Millipore, Billerica, MA, USA). After blocking with 5% skim milk in Tris-buffered saline (pH 7.4) containing 0.05 % Tween-20 (TBS-T) for 1 h at room temperature (RT), the membrane was incubated with primary antibodies overnight at 4°C followed by incubation with horseradish peroxidase-conjugated secondary antibodies (Jackson ImmunoResearch Lab., Inc., West Grove, PA, USA) for 1 h at RT or overnight at 4 °C. The membrane was reacted with ECL (GE Healthcare, Uppsala, Sweden) for the detection of -actin, G-protein-coupled receptor kinase (GRK) 2, and total a-Syn, including phosphorylated and non-phosphorylated forms, or ECL plus (GE Healthcare) for the detection of other proteins. The membrane was then visualized using a CCD camera, VersaDog 5000 (Bio-Rad, Hercules, CA, USA). Relative intensities of detected signals were quantified with Quantity one software (Bio-Rad). For detection of phosphorylated a-Syn, we added 50 mM NaF into TBS-T containing 5% milk or antibodies.

The effect of protein phosphatase 2A/1 inhibitor on the levels of Ser129-phosphorylated a-Syn in SH-SY5Y cells.
The metabolic fate of a phosphorylated protein is affected by dephosphorylation and/or degradation pathways, such as the proteasome or autophagy-lysosome pathway. Fujiwara et al. has reported that Ser129-phosphorylated a-Syn protein underwent dephosphorylation (7). This idea is supported by the findings showing that the protein phosphatase 2A/1 inhibitor, okadaic acid, increases the amount of phosphorylated a-Syn in PC12 cells (15) and HEK293 cells (11). To compare the contribution of dephosphorylation to the metabolism of Ser129-phosphorylated a-Syn with that of degradation pathways, we first assessed the effect of okadaic acid on the metabolism of Ser129-phosphorylated a-Syn. In this study, we used a cell line (wt-aS/SH) stably expressing wild-type a-Syn, because it is difficult to detect Ser129-phosphorylated a-Syn at endogenous levels in SH-SY5Y cells. As compared with the starting levels, the expression levels of Ser129 phosphorylated a-Syn were increased 2.60 ± 0.43-fold at 2 h (mean ± SD, p < 0.01, n = 4) and 5.53 ± 2.32-fold at 4 h (p = 0.03, n = 4) in the presence of 10 nM okadaic acid (Fig. 1). In contrast, the expression levels of total a-Syn, including non-phosphorylated and phosphorylated forms, were constant at 0.92 ± 0.04-fold at 2 h (p = 0.125, n = 3) and 1.03 ± 0.23-fold at 4 h (p = 0.861, n = 3), when compared with starting levels (Fig. 1). Unexpectedly, the expression levels of Ser129-phosphorylated a-Syn were moderately but significantly increased 1.82 ± 0.44-fold at 4 h in the presence of 0.1% DMSO (p = 0.014, n = 5) (Fig. 1). When the cells were incubated in medium containing 0.1% PBS for 4 h, the levels of Ser129-phosphorylated a-Syn were constant (Fig. 1). The increased effect of okadaic acid on the levels of phosphorylated a-Syn was larger than that of DMSO at 4 h post-treatment (5.53 ± 2.32-fold in okadaic acid versus 1.82 ± 0.44-fold in DMSO) (Fig. 1).
The effect of proteasome inhibitors on the levels of Ser129-phosphorylated a-Syn in SH-SY5Y cells.
The effect of autophagy-lysosome inhibitors on the levels of Ser129-phosphorylated a-Syn in SH-SY5Y cells.
We next tested whether Ser129-phosphorylated a-Syn was degraded by the autophagy-lysosome pathway. As compared with the starting levels, the expression levels of Ser129-phosphorylated a-Syn were increased 2.05 ± 0.52-fold at 32 h in the presence of 10 mM 3-MA, which inhibits the formation of the autophagosome (p = 0.027, n = 3) (Fig. 3). The expression levels of total a-Syn were also increased 1.5 ± 0.06-fold at 32 h (p = 0.005, n = 3) (Fig. 3). Additionally, treatment with 100 μM of general lysosomal inhibitor, chloroquine, showed that the expression levels of Ser129-phosphorylated a-Syn were 6.52 ± 2.20-fold higher than the starting levels at 32 h (p = 0.049, n = 3) (Fig. 3). The expression levels of total a-Syn were found to be increased 1.39 ± 0.10-fold at 32 h (p = 0.022, n = 3) (Fig. 3). When the cells were treated with 0.1% PBS as a vehicle control, the levels of Ser129-phosphorylated and total a-Syn were constant. Inhibition of the autophagy-lysosome pathway showed that the expression levels of Ser129-phosphorylated a-Syn were elevated almost in parallel to those of total a-Syn by longer incubations than inhibition of the proteasome pathway.
Role of dephosphorylation and proteasomal degradation in the metabolism of Ser129-phosphorylated a-Syn in SH-SY5Y cells.
To elucidate the contribution of dephosphorylation and degradation pathways to the metabolism of Ser129-phosphorylated a-Syn, we determined the half-life of Ser129-phosphorylated a-Syn by using the de novo protein synthesis inhibitor, cycloheximide, in wt-aS/SH cells. As shown in Figure 4A and B, the experiment using cycloheximide showed that the amounts of Ser129-phosphorylated a-Syn were rapidly decreased. Its estimated half-life (t 1/2 ) was 54.9 ± 6.4 min. In contrast, the expression levels of total a-Syn were stable within the observation time of up to 240 min (Figs. 4A&B). Because this finding suggested that Ser129-phosphorylated a-Syn was rapidly processed by the specific pathway that differed from the non-phosphorylated form, we then focused on the role of dephosphorylation and degradation by the proteasome pathway. We treated wt-aS/SH cells with cycloheximide plus either okadaic acid or MG132. To ensure that okadaic acid or MG132 exerts its inhibitory effect during the entire course of the experiment, we pre-treated the cells with the reagents until accumulation of Ser129-phosphorylated a-Syn was easily detectable. Under these conditions, treatment with okadaic acid suppressed the decrease in the amounts of Ser129- The amounts of total a-Syn did not change in the presence of okadaic acid or MG132 (Figs. 4A&B). These data suggested that Ser129-phosphorylated a-Syn underwent degradation mainly by the proteasome pathway rather than dephosphorylation in SH-SY5Y cells. We performed pulse-chase experiments after metabolic labeling. Ser129-phoshorylated a-Syn was decreased between 4 and 6 h of the chase periods, while total a-Syn was almost stable in the chase period (Fig. 4C). This finding was consistent with the data from experiments using cycloheximide.
To exclude a possibility that the decrease in Ser129-phosphorylated a-Syn was due to degradation of kinases, we examined whether treatment with cycloheximide altered the expression levels of GRKs and CK2, which have been known to contribute to the Ser129-phosphorylation of a-Syn (11,13,15,18,19). The expression levels of ubiquitously expressing members of GRKs (GRK2, -3, -5, and -6) and CK2 α'-subunit were stable in the presence of cycloheximide and in the presence of cycloheximide plus either okadaic acid or MG132 during the observation times (Fig. 4D).
We next assessed the metabolism of phosphorylation-mimic mutants (Ser129Glu and Ser129Asp) and phosphorylation-abolished mutant (Ser129Ala) of a-Syn. Each mutant cDNA was transfected into SH-SY5Y cells, and the cells were treated with lactacystin for 4 h. The expression levels of Ser129Ala mutant a-Syn were stable at 4 h, as compared with the starting levels (Fig. 5A). In phosphorylation-mimic mutants, the expression levels of Ser129Glu mutant a-Syn were significantly increased 1.39 ± 0.15-fold at 2 h (p = 0.014, n = 4) and 1.55 ± 0.23-fold at 4 h (p = 0.016, n = 4), as compared with the starting levels, whereas Ser129Asp mutant a-Syn did not alter the expression levels (Fig. 5A). In the experiments using cycloheximide, Ser129Ala and Ser129Asp mutants were not decreased within 4 h (Fig. 5B). On the other hand, Ser129Glu mutant a-Syn showed a rapid decrease (Fig. 5B). Treatment with MG132 suppressed the rapid decrease in the levels of Ser129Glu mutant a-Syn (Fig. 5B). The data of Ser129Ala and Ser129Glu mutants supported that Ser129-phosphorylation played a role in the degradation of a-Syn by the proteasome pathway. However, Ser129Asp mutant a-Syn did not reproduce the metabolic fate of Ser129-phosphorylated form.
To test whether the phosphorylation of other Ser residues contributed to the metabolism of Ser129-phosphorylated a-Syn, we made Ser9Ala/Ser42Ala/Ser87Ala mutant a-Syn cDNA, whose product abolished the possible Ser-phosphorylation sites except for Ser129. In the experiment using cycloheximide, Ser129-phosphorylated form of this mutant a-Syn was rapidly decreased (Fig. 6). The expression levels of total a-Syn in the mutant were stable (Fig.  6). Treatment with MG132 suppressed the rapid decrease in the levels of Ser129-phosphorylated form of the mutant a-Syn (Fig. 6).

Effects of dephosphorylation and the proteasome pathway on the metabolism of Ser129-phosphorylated a-Syn in rat primary cortical neurons.
To assess whether or not overexpression of a-Syn artificially targeted Ser129-phosphorylated a-Syn towards degradation by the proteasome pathway, we investigated the effect of the proteasome inhibitors on the endogenous a-Syn protein in rat primary cortical neurons. When 21 day cultured neurons were incubated in the presence of proteasome inhibitors for 4 h, the expression levels of Ser129-phosphorylated a-Syn were increased 2.84 ± 0.59-fold in the presence of lactacystin (p = 0.008, n = 4) and 2.03 ± 0.39-fold in the presence of MG132 (p = 0.013, n = 4), as compared with starting levels (Fig. 7). The expression levels of total a-Syn were constant in the presence of lactacystin (1.04 ± 0.17-fold, p = 0.637, n = 4) and MG132 (1.07 ± 0.30-fold, p = 0.730, n = 4) (Fig.  7). Additionally, when the primary neurons were treated with okadaic acid, the expression levels of Ser129-phosphorylated a-Syn were increased 2.40 ± 0.69-fold (p = 0.007, n = 4), as compared with starting levels (Fig. 7). Although treatment with DMSO showed a tendency to elevate the levels of phosphorylated a-Syn, there was no significant difference between 0 and 4 h. The data demonstrated that Ser129-phosphorylated a-Syn underwent degradation by the proteasome pathway and dephosphorylation at the endogenous level in neurons.

Role of ubiquitination in the degradation of Ser129-phosphorylated a-Syn by the proteasome pathway.
To assess whether poly-ubiquitination plays a role in targeting Ser129-phosphorylated a-Syn to the proteasome pathway, we investigated wt-aS/SH cells and parental SH-SY5Y cells in the absence or presence of lactacystin for 16 h. Immunoblotting of post-12,000  g centrifuged supernatant fractions with anti-ubiquitin antibody showed that poly-ubiquitin conjugates were increased in the presence of lactacystin in both wt-aS/SH and parental cells (Fig. 8A). In post-12,000  g centrifuged pellet fractions, poly-ubiquitin conjugates were also increased in the presence of lactacystin in these cells (Fig. 8A). These findings indicated that lactacystin treatment effectively blocked protein degradation by the proteasome pathway, resulting in accumulation of poly-ubiquitinated conjugates in the cells. However, there was no difference in the pattern of ubiquitin-positive bands between wt-aS/SH cells and parental cells in both supernatant and pellet fractions (Fig. 8A). In supernatant fractions of lactacystin-treated wt-aS/SH cells, immunoblotting with anti-Ser129-phosphorylated a-Syn antibody showed that monomeric Ser129-phosphorylated a-Syn was increased; however, there was no appearance of the specific bands migrating at a position of higher molecular weight than its monomer (Fig. 8A). Although we analyzed the pellet fractions of wt-aS/SH cells, the specific bands corresponding to poly-ubiquitinated Ser129-phosphorylated a-Syn were not detectable (Fig. 8A). We also found no specific band in immunoblotting with anti-total a-Syn antibody (Fig. 8A). We next investigated wt-aS/SH cells transiently overexpressing ubiquitin. Although poly-ubiquitinated conjugates were clearly increased in supernatant and pellet fractions by treatment with lactacystin, there was no appearance of the specific bands corresponding to poly-ubiquitinated a-Syn in the immunoblots of these fractions with anti-ubiquitin, anti-Ser129 phosphorylated a-Syn or anti-total a-Syn antibody (Fig. 8A). Additionally, we did not find poly-ubiquitinated a-Syn in rat primary cortical neurons in the presence of MG132 or lactacystin, as compared with the neurons in the absence of inhibitor (Fig. 8A).
We further examined ubiquitination of a-Syn by immunoprecipitation of wt-aS/SH and parental cells transfected with FLAG-tagged ubiquitin cDNA (Fig. 8B). In products immunoprecipitated with anti-total a-Syn antibody, we detected the a-Syn monomer by immunoblotting with anti-total a-Syn antibody (Fig. 8B). However, there was no signal showing poly-ubiquitination of a-Syn in the immunoprecipitated products of wt-aS/SH cells in the absence or presence of MG132 (Fig. 8B). Additionally, in products immunoprecipitated with anti-FLAG antibody, we detected poly-ubiquitinated conjugates by immunoblotting with anti-ubiquitin antibody (Fig. 8B). However, there was no signal showing poly-ubiquitination of a-Syn in the immunoprecipitated products of wt-aS/SH cells in the absence or presence of MG132 (Fig. 8B).
To clarify whether the present biochemical fractionation experiments failed to detect insoluble ubiquitinated a-Syn by the aggregate formation, we performed immunocytochemical analysis using wt-aS/SH cells. The cells were transiently transfected with ubiquitin cDNA and treated with 10 μM MG132 for 16 h. Overexpressed a-Syn and ubiquitin proteins were diffusely distributed in the cytoplasm in the absence or presence of MG132 (Fig. 9A). The formation of a-Syn-or ubiquitin-positive inclusions was not found in the presence of MG132 (Fig. 9A). These findings were consistent with the biochemical fractionation data that there were no ubiquitinated a-Syn proteins in the pellet fractions under the present condition. Also, the formation of a-Syn-or ubiquitin-positive inclusions was not detected in the presence of okadaic acid (Fig. 9A).
To elucidate whether ubiquitination of lysine residues of a-Syn contributed to the metabolism of Ser129-phosphorylated a-Syn, we made a Lys12Arg/Lys21Arg/Lys23Arg mutant a-Syn cDNA, whose product abolished the previously reported ubiquitination sites of a-Syn (8). In the experiment using cycloheximide, Ser129-phosphorylated form of this mutant a-Syn was rapidly decreased (Fig. 9B). The expression levels of total a-Syn in the mutant were constant (Fig. 9B). Treatment with MG132 suppressed the rapid decrease in the levels of Ser129-phosphorylated form of the mutant a-Syn (Fig. 9B).

DISCUSSION
The present data demonstrated that Ser129-phosphorylated a-Syn underwent dephosphorylation and degradation. In SH-SY5Y cells stably expressing a-Syn, inhibition of the proteasome pathway or the autophagy-lysosome pathway resulted in the accumulation of Ser129-phosphorylated a-Syn. However, the inhibitory effect of the proteasome pathway was different in two points from that of the autophagy-lysosome pathway. First, inhibition of the proteasome pathway did not accompany the alteration in the levels of total a-Syn. Second, inhibition of the proteasome pathway increased the levels of Ser129-phosphorylated a-Syn faster than that of the autophagy-lysosome pathway. To elucidate these differences, we assessed the half-life of Ser129-phosphorylated a-Syn using cycloheximide. The result showed that the half-life of Ser129-phosphorylated a-Syn was much shorter than that of total a-Syn. Inhibition of the proteasome pathway remarkably prolonged the short half-life of Ser129-phosphorylated a-Syn with no alteration in that of total a-Syn. In rat primary cortical neurons, inhibition of the proteasome pathway also accumulated phosphorylated a-Syn at endogenous levels. These findings suggest that Ser129-phosphorylated a-Syn specifically undergoes degradation by the proteasome pathway, and that the portion of Ser129-phosphorylated a-Syn is too small to affect the levels of total a-Syn. On the other hand, contribution of the autophagy-lysosome pathway to degradation of phosphorylated a-Syn remains to be elucidated. One may speculate that the autophagy-lysosome pathway selectively degrades non-phosphorylated a-Syn. The increase in the levels of phosphorylated a-Syn may be a consequence of accumulation of non-phosphorylated proteins in the cytosol due to inhibition of the autophagy-lysosome pathway. Alternatively, the autophagy-lysosome pathway may degrade both non-phosphorylated and phosphorylated a-Syn. In this case, the portion of phosphorylated a-Syn is estimated to be very small, because this corresponds to phosphorylated a-Syn, which remains undiminished in the present experiments using cycloheximide.
Since Bennett et al. initially reported that inhibition of the proteasome pathway led to the accumulation of overexpressed a-Syn in SH-SY5Y cells (20), this finding has also been shown in more recent studies (21)(22)(23)(24). However, other studies have failed to detect accumulation of endogenous or overexpressed a-Syn by inhibition of the proteasome pathway (25)(26)(27)(28). In contrast, a-Syn is reported to accumulate in cells by inhibition of the autophagy-lysosome pathway (29)(30)(31). At present, the involvement of the proteasome pathway or the autophagy-lysosome pathway in degradation of a-Syn is still debated. In general, proteins with short half-lives are mostly degraded by the proteasome pathway, whereas most cytosolic proteins with long half-lives (> 10 h) are degraded by the lysosome pathway (30). In contrast to this principle, a previous study utilized long incubation times over 48 h to detect accumulation of a-Syn by inhibition of the proteasome pathway (21). Additionally, treatment with a selective proteasome inhibitor, epoxomicin, only showed a 2.3 h increase in the half-life of a-Syn (16.8 ± 2 h) in rat ventral midbrain cultures (30). The findings suggest that a portion of a-Syn, which is targeted to the proteasome pathway, is small. These studies may have seen accumulation of Ser129-phosphorylated a-Syn by inhibition of the proteasome pathway. Also, the present data were consistent with the finding that a-Syn was degraded by the autophagy-lysosome pathway. Ser129-phosphorylation may be a key for resolving whether the metabolism of a-Syn fulfills this general principle.
In the present study, we investigated the role of Ser residues in targeting a-Syn to the proteasome pathway. a-Syn has four Ser residues and is known to be phosphorylated at Ser87, as well as at Ser129 (15). To assess the effect of Ser residues other than Ser129 on proteasomal degradation, we constructed the Ser9Ala/Ser42Ala/Ser87Ala mutant of a-Syn. In experiments using cycloheximide, overexpressed mutant a-Syn demonstrated a rapid decrease in the levels of the Ser129-phosphorylated form and the stable expression of total protein. The rapid decrease in the levels of the Ser129-phosphorylated form was inhibited by MG132. These findings were similar to wild-type a-Syn, suggesting that Ser129 may play a central role in targeting the protein to the proteasome pathway. However, this study could not exclude a possibility that targeting of a-Syn to the proteasome pathway depends on the phosphorylation of other residues, because a-Syn is also known to be phosphorylated at three tyrosine residues.
We also assessed the metabolism of phosphorylation-mimic (Ser129Glu  and  Ser129Asp) and phosphorylation-abolished (Ser129Ala) mutants of a-Syn. Ser129Ala mutant a-Syn did not show the degradation through the proteasome pathway. In contrast, Ser129Glu mutant a-Syn was degraded by the proteasome pathway. These findings were consistent with the proteasomal degradation of Ser129-phosphorylated a-Syn. However, Ser129Asp mutant a-Syn was not targeted to the proteasome pathway. This finding indicated that substitution of aspartic acid for Ser129 did not mimic the phosphorylated residue. The previous paper reported that phosphorylation-mimic mutants of a-Syn did not reproduce the effect of the phosphorylation on the structural properties of a-Syn in vitro (10). One may speculate that conformation changes induced by the phosphate group, rather than negative charge, are responsible for the effect of phosphorylation on the metabolism of a-Syn. Further studies are required to elucidate the determinants that lead to the difference in the metabolism between Ser129Glu and Ser129Asp mutants of a-Syn.
It is well established that conjugation of at least four ubiquitins on a protein is necessary for ubiquitin-dependent degradation to occur (32). However, previous studies have shown that ubiquitination of unmodified a-Syn does not occur after inhibition of the proteasome in transfected cells (20,27,28). In vitro experiments using purified recombinant proteins have shown that unmodified a-Syn is degraded by the 20S (21,23,24) and 26S proteasomes (23). These findings indicate that a-Syn is degraded by the proteasome pathway in a ubiquitin-independent manner (21). Although the present data support the previous findings, we propose an idea that Ser129-phosphorylated a-Syn is targeted to the proteasome pathway in a ubiquitin-independent manner.
What is the role of Ser129-phosphorylation in the proteasomal degradation of a-Syn? The physiological function of a-Syn is thought to require its association with lipid vesicles where it adopts an α-helical conformation from a natively unstructured one (33). Structural studies have shown that a-Syn dynamically binds to vesicles and is promoted to dissociate from the vesicle by mutation (34), oxidative stress (35) and Ser129-phosphorylation (18). Liu et al. reported that 20S proteasome degraded free, unstructured a-Syn, but not vesicle-bound, α-helical a-Syn (24). The Ser129-phosphorylation may trigger the dissociation of a-Syn proteins from vesicles and accumulate free, unstructured proteins in the cytosol, resulting in targeting them to the proteasome pathway.
In summary, our data suggest that Ser129-phosphorylation plays a role in the rapid degradation of a-Syn by targeting the protein to the proteasome pathway. Bedford et al. reported that depletion of the 26S proteasome in mouse neurons caused extensive neurodegeneration in the nigrostriatal pathway and the formation of LB-like inclusions containing a-Syn (36). This study strongly indicates that 26S proteasome dysfunction in neurons is involved in aggregation of a-Syn (36). Ser129-phosphorylation might provide a clue for linking the degradation pathway of a-Syn with the formation of LBs. However, the present study does not resolve the question why Ser129-phosphorylated a-Syn proteins deposited in LBs are also ubiquitinated (9,37). Our data suggest that phosphorylated a-Syn proteins undergo ubiquitination in the pathway independent of its physiological degradation. Ubiquitination might represent an unsuccessful "last-ditch stand" of cells in their attempt to unfold and/or degrade misfolded proteins in a disease-specific pathway (37). Alternatively, ubiquitination of a-Syn might occur after polymerization of the molecule (38). Further studies are required to clarify whether the proteasome dysfunction accumulates Ser129-phosphorylated a-Syn selectively, resulting in generation of LBs, which abundantly contain Ser129-phosphorylated a-Syn. B. Quantitative analysis of the alteration in the expression levels of Ser129 phosphorylated a-Syn in cells in the absence or presence of okadaic acid. Relative ratios of the band intensity of Ser129-phosphorylated a-Syn to β-actin and total a-Syn to β-actin are shown. Relative ratio was normalized to the starting material just before adding reagents. Data represent means ± S.D. and the p-values (*p<0.05, **p<0.01) were estimated by unpaired Student's t-test.  Relative ratios of the band intensity of Ser129-phosphorylated a-Syn to β-actin and total a-Syn to β-actin are shown. Relative ratio was normalized to the starting material. Data represent means ± S.D. and the p-values (*p<0.05, **p<0.01) are shown.