Assembly of Lysine 63-linked Ubiquitin Conjugates by Phosphorylated α-Synuclein Implies Lewy Body Biogenesis*

α-Synuclein (α-syn) and ubiquitin (Ub) are major protein components deposited in Lewy bodies (LBs) and Lewy neurites, which are pathologic hallmarks of idiopathic Parkinson disease (PD). Almost 90% of α-syn in LBs is phosphorylated at serine 129 (Ser129). However, the role of Ser129-phosphorylated α-syn in the biogenesis of LBs remains unclear. Here, we show that compared with coexpression of wild type (WT)α-syn and Ub, coexpression of phospho-mimic mutant α-syn (S129D) and Ub in neuro2a cells results in an increase of Ub-conjugates and the formation of ubiquitinated inclusions. Furthermore, S129D α-syn fails to increase the Ub-conjugates and form ubiquitinated inclusions in the presence of a K63R mutant Ub. In addition, as compared with WT α-syn, S129D α-syn increased cytoplasmic and neuritic aggregates of itself in neuro2a cells treated with H2O2 and serum deprivation. These results suggest that the contribution of Ser129-phosphorylated α-syn to the Lys63-linked Ub-conjugates and aggregation of itself may be involved in the biogenesis of LBs in Parkinson disease and other related synucleinopathies.

Parkinson disease (PD) 2 is the most common neurodegenerative movement disorder (1). Pathologically, it is characterized by loss of dopamine neurons and the presence of cytoplasmic inclusions (Lewy bodies, LBs) in surviving neurons in the substantia nigra pars compacta, accompanied by the presence of dystrophic neurites (Lewy neurites) (1,2). Although many studies have focused on the LBs, the mechanism that underlies LB biogenesis is poorly understood (3,4).
Ubiquitination is a process whereby a small protein called ubiquitin (Ub) conjugates to its target protein. Monoubiquitination occurs through the isopeptide bond between the C-terminal glycine (Gly-76) residue on Ub and the ⑀-amino group of the lysine (Lys) side chain in the target protein by multisteps that are activated sequentially by E1 (Ub-activating), E2 (Ubconjugating), and E3 (Ub-ligase) enzymes (19). The addition of one or more Ub moieties to the Ub on the target protein results in di-or polyubiquitination. Ub, one of the major components in LBs, is an essentially required component in the ubiquitinproteasome system (UPS) (20,21). UPS is an intracellular proteolytic system that degrades its targeted proteins (22,23). To date, many studies suggest that malfunction of protein degradation linked to UPS plays an important role in the pathogenesis of PD (23,24). A protein tagged by a chain of four or more Ubs, instead of a single Ub, is targeted for proteasomal degradation (25). Specifically, Lys 48 -linked poly-Ub chains instead of Lys 63 -linked poly-Ub chains are considered to be a proteasomal degradation signal (22,23). In vivo, the aggregation of ␣-syn is enhanced by Lys 63 -linked ubiquitination (26). Also, both wild type ␣-syn and mutant ␣-syn, incubated with rabbit reticulocyte Fraction IIA in vitro, increase Lys 63 -linked Ub-conjugate formation with the mutants being more effective (27). The increased Ub-conjugates are detectable by anti-Ub antibodies, but not by anti-␣-syn antibody, indicating that ␣-syn may function in regulating assembly of Ub-Lys 63 chains (27).
In this study, we investigate the possible role of phosphorylation of ␣-syn at Ser 129 in the biogenesis of LBs using a cellular model. Coexpression of S129D ␣-syn, on which serine 129 was converted to aspartate to mimic phosphorylated ␣-syn, and Ub resulted in an increase in Ub-conjugates and cytoplasmic ubiquitinated aggregates compared with the coexpression of WT ␣-syn with Ub. We also observed that coexpression of S129D ␣-syn and Ub caused the formation of ubiquitinated inclusions in neuro2a cells. Furthermore, in neuro2a cells cotransfected with K63R mutant Ub, S129D ␣-syn failed to increase the Ubconjugates or form ubiquitinated inclusions. Compared with WT ␣-syn, S129D ␣-syn increased the ubiquitination of itself when coexpressed with Ub. Additionally, S129D ␣-syn formed more aggregates in neuro2a cells treated with H 2 O 2 and serum deprivation than WT ␣-syn. Our data may provide insight into a molecular mechanism that links phosphorylation of ␣-syn and Lys 63 -linked Ub-conjugates to LB biogenesis.
Cell Culture and Transfection-Neuro2a cells cultured overnight in Dulbecco's modified Eagle's medium (DMEM) (Invitrogen) containing 10% newborn calf serum (Invitrogen) were washed with Opti-MEM and then transiently transfected with expressing plasmids using Lipofectamine TM 2000 reagent (Invitrogen) in Opti-MEM without serum. The same volume of DMEM containing 10% newborn calf serum was added to the culture media 6 h after transfection. Two days later, the transfected cells were observed by an inverted microscope (Olym-pus, IX71, Japan) or subjected to fluorescent observation, immunoblot, or coimmunoprecipitation analysis.
Immunocytochemistry-Transfected cells grown in chamber slides were washed with PBS, fixed in PBS with 4% paraformaldehyde for 5 min, and then treated with 0.25% Triton X-100 for 10 min. After blocking in 4% BSA/PBS for 1 h, the cells were incubated with anti-FLAG antibody (1:5000, Sigma) for 1 h at room temperature. After washing with PBS, the cells were incubated with rhodamine-conjugated donkey anti-mouse antibodies (1:200, Santa Cruz Biotechnology) for 1 h at room temperature. Then the cells were washed with PBS and incubated with DAPI (Sigma) for 3 min at room temperature. Finally, the cells were observed using an inverted fluorescence microscope.
Coimmunoprecipitation-Transfected cells were treated with 10 M MG132 (Calbiochem, La Jolla, CA) for 12 h, 48 h after transfection. The cells were harvested and sonicated in modified TSPI buffer (50 mM Tris-HCl (pH 7.5), 150 mM sodium chloride, 1 mM EDTA, 1 g/ml of aprotinin, 10 g/ml of leupeptin, 0.5 M Pefabloc SC, 10 g/ml of pepstatin, and 1 mM phenylmethylsulfonyl fluoride) containing 1% Nonidet P-40. Cellular debris was removed by centrifugation at 16,000 ϫ g for 30 min at 4°C. The supernatants were preincubated with protein G-Sepharose (Roche) for 2 h at 4°C and then incubated with monoclonal anti-GFP antibody (Roche) for 1 h at 4°C. After incubation, protein G-Sepharose was used for precipitation. The beads were washed with TSPI buffer four times and then eluted with SDS sample buffer for immunoblot analysis.
Fractionation Experiments-For Nonidet P-40 soluble or insoluble fractionation experiments, cells were lysed in a modified TSPI buffer. After sonication, cells were centrifuged at 16,000 ϫ g for 30 min at 4°C. Nonidet P-40-insoluble pellets were dissolved in a buffer containing 1% SDS, 1% Nonidet P-40. The soluble and insoluble fractions (dissolved by 1% SDS) were subjected to immunoblot analysis using the antibodies previously described.
Aggregation Induction-Cells were transfected with various constructs. At 24 h after transfection, the DMEM containing 10% newborn calf serum was replaced with newborn calf serum-free DMEM containing 400 M H 2 O 2 . Cells were cultured for a further 72 h, then washed with PBS, fixed in PBS with 4% paraformaldehyde for 5 min, and then treated with 0.25% Triton X-100 for 10 min at room temperature. After being washed with PBS, cells were incubated with DAPI (Sigma) for 3 min at 4°C. After incubation, cells were washed with PBS and observed using an inverted fluorescence microscope.

RESULTS
Characterization of WT ␣-Syn, S129D ␣-Syn, or ␤-Syn Expressed in Neuro2a Cells-It has been reported that almost 90% of ␣-syn in LBs is phosphorylated at Ser 129 (13). To investigate the possible role of Ser 129 -phosphorylated ␣-syn in LB biogenesis, we created a mutant form of ␣-syn in which serine (S) 129 was converted to aspartate (D) to mimic phosphorylated ␣-syn (S129D) (Fig. 1A). To use enhanced green fluorescent protein (EGFP) as a reporter, we generated plasmids expressing EGFP-tagged WT ␣-syn, S129D ␣-syn or ␤-syn (Fig. 1A). To exclude the possible effects of EGFP on the ␣-syn conformation, we also created plasmids expressing V5-tagged WT ␣-syn or S129D ␣-syn (Fig. 1A, bottom). Neuro2a cells were transfected with EGFP, EGFP-tagged WT ␣-syn, S129D ␣-syn or ␤-syn. Cells overexpressing these EGFP-tagged proteins dis-played a diffusive distribution in cytoplasm, nuclei, and neurites ( Fig. 1, B-E). Forty-eight hours after transfection, cells were harvested and the total cell lysates were subjected to immunoblot analysis using anti-GFP antibody. Expression of all the constructs was confirmed by immunoblot analysis (Fig. 1J). Aggregation of these fusion proteins was not observed in transfected neuro2a cells. Transfection with V5-lacZ, V5-tagged WT, or S129D ␣-syn was also performed. Forty-eight hours after transfection, cells were harvested and the soluble and insoluble fractions dissolved with 1% SDS were subjected to immunoblot analysis using anti-V5 antibody (Fig. 1K). Aggregation of these fusion proteins was not observed in transfected neuro2a cells.
Coexpression of S129D ␣-Syn and Ub Increases Ub-conjugates and Induces LB-like Structure Formation-To investigate whether the phosphorylation of ␣-syn at serine 129, which has been linked to the pathogenesis of PD, has effects on UPS, we examined the ubiquitinated proteins in neuro2a cells cotransfected with FLAG-Ub along with EGFP, EGFP-tagged WT ␣-syn or S129D ␣-syn. We also examined cells cotransfected with FLAG-Ub along with V5-lacZ, V5-tagged WT ␣-syn, or S129D ␣-syn. Forty-eight hours after transfection, the cells were analyzed by immunoblotting or immunofluorescence. Immunoblotting analysis revealed the smear bands of anti-FLAG immunoreactivity that represented the Ub-conjugates. In Nonidet P-40-soluble fractions, coexpression of WT or S129D ␣-syn with Ub resulted in more Ub-conjugates, especially in the high molecular weight, than coexpression of Ub and EGFP or V5-lacZ (Fig. 2, A and B). In Nonidet P-40-insoluble fractions, S129D ␣-syn resulted in more Ub-conjugates than WT ␣-syn (Fig. 2, A and B). As a comparison, the S129A ␣-syn mutant, where Ser 129 was converted to alanine (A) to abolish phosphorylation at this site, was used. Coexpression of S129A ␣-syn and FLAG-Ub did not contribute to the Ub-conjugate formation compared with coexpression of FLAG-Ub and WT or S129D ␣-syn (supplemental Fig. S1). However, when using anti-GFP or anti-␣-syn antibody, no low migrating bands, compatible with the ubiquitinated ␣-syn species, were detected in cell lysates (data not shown). The transfected cells were visualized by excitation of EGFP and immunochemistry analysis using the anti-FLAG antibody. As shown in Fig. 2, C-K, the number of ubiquitinated aggregates per Ub-positive cell was counted and subjected to data analysis. About 50% of Ub-positive cells, in the presence of EGFP and Ub, formed ubiquitinated aggregates. In the presence of WT ␣-syn and Ub, ubiquitinated aggregates were found in more than 60% of Ub-positive cells. Whereas in the presence of S129D ␣-syn and Ub, cells with ubiquitinated aggregates increased significantly to ϳ80% of Ub-positive cells (Fig. 2L). The quantitative data showed that coexpression of S129D ␣-syn and Ub significantly increased the number of cells with three or more ubiquitinated aggregates as compared with coexpression of Ub and EGFP or WT ␣-syn (Fig. 2M). These results are consistent with our immunoblotting results suggesting that phosphorylated ␣-syn contributes to the accumulation of ubiquitinated proteins.
Characterization of LB-like Structure Biogenesis Induced by Coexpression of S129D ␣-Syn and Ub-Neuro2a cells were cotransfected with EGFP-tagged S129D ␣-syn and FLAG-Ub. Forty-eight hours after transfection, EGFP-tagged S129D ␣-syn FIGURE 1. Characterization of synucleins. A, schematic diagram of constructs encoding EGFP-tagged ␣-syn or ␤-syn and V5-tagged ␣-syn. The serine 129 residue is shown near the C terminus of ␣-syn. B-I, neuro2a cells were transfected with EGFP or EGFP-tagged with WT ␣-syn, S129D ␣-syn or ␤-syn. Cells were observed using an inverted fluorescence microscope (Olympus, IX71, Japan). EGFP and EGFP-tagged proteins are shown in green (B-E). Cell number of each sample is shown by phase observation (F-I). Bar, 10 m. J, expression of EGFP-tagged constructs was examined by immunoblot analysis using anti-GFP antibody. K, expression of V5-tagged constructs was examined by immunoblot analysis using anti-V5 antibody. WB, Western blot; S, soluble; IS, insoluble.
was visualized by excitation of EGFP and FLAG-Ub was visualized by immunocytochemistry analysis using anti-FLAG antibody. There are different patterns of ␣-syn intracytoplasmic aggregation in cotransfected cells (Fig. 3, A-T). Some cells contained diffusive or "cloud-like" ␣-syn fluorescence (Fig. 3,  A-D). Occasionally, this diffuse ␣-syn fluorescence showed a more intense aggregation or a greater tendency to stain an outer rim, whereas Ub was more likely to stain the core (Fig. 3, E-H). Colocalization of ␣-syn with Ub was found in both large and small aggregates scattered in the cytoplasm (Fig. 3, I-L). Two patterns of ubiquitinated inclusions were observed in neuro2a cells: one shows amorphous inclusion with Ub immunoreactivity and ␣-syn fluorescence (Fig. 3, M-P); another shows well defined colocalization of ␣-syn with Ub (Fig. 3, Q-T).
Effects of S129D ␣-Syn on Ub-conjugates and LB-like Structure Formation Depend on Lysine 63 on Ub-Ubiquitination of specific proteins that target themselves to proteasomal degradation is believed to be dependent on the conjugation of the poly-Ub chain, which is linked through specific lysines on Ub (22). The most prominent sites are the degradation related Lys 48 and degradation unrelated Lys 63 on Ub. To further investigate whether Lys 48 or Lys 63 on Ub is involved in the Ubconjugate formation and why these conjugates escape from proteasomal degradation, we created two mutant forms of Ub on which lysine 48 or 63 was converted to arginine. We examined the ubiquitinated proteins in neuro2a cells cotransfected with K48R or K63R mutant FLAG-Ub together with EGFP, EGFP-tagged WT ␣-syn, or S129D ␣-syn for 48 h. Immunoblotting analysis using anti-FLAG antibody showed that coexpression of WT or S129D ␣-syn with the K48R mutant Ub both resulted in more poly-Ubconjugates, especially in the high molecular weight, as compared with coexpression of EGFP and K48R mutant Ub (Fig. 4A). Furthermore, in Nonidet P-40 insoluble fractions, S129D ␣-syn resulted in more Ubconjugates than WT ␣-syn. These results are similar to our observations in cells that were cotransfected with WT or S129D ␣-syn and wild type Ub ( Fig. 2A). Coexpression of WT or S129D ␣-syn with K63R mutant Ub did not contribute to poly-Ub-conjugate formation in either Nonidet P-40-soluble or -insoluble fractions any more than coexpression of EGFP with K63R mutant Ub (Fig. 4B). To further ascertain whether Lys 63 on Ub is involved in LB-like structure formation, neuro2a cells were cotransfected with K63R mutant FIGURE 2. S129D ␣-syn promotes Ub-conjugates and ubiquitinated inclusion formation. A, neuro2a cells were cotransfected with FLAG-Ub along with EGFP, EGFP-tagged WT ␣-syn, or S129D ␣-syn for 48 h. B, neuro2a cells were cotransfected with FLAG-Ub along with V5-lacZ, V5-tagged WT ␣-syn, or S129D ␣-syn for 48 h. Cell lysates were subjected to fractionation experiments. Both soluble and insoluble fractions were subjected to immunoblot analysis using anti-FLAG, anti-GFP, or anti-V5 antibody. In both the soluble and insoluble fractions, the loading amount of each sample was examined by anti-␣-tubulin immunoblot analysis. FLAG-Ub along with EGFP, EGFP-tagged WT ␣-syn, or S129D ␣-syn. Forty-eight hours after transfection, the cells were visualized by excitation of EGFP and by immunochemistry analysis using anti-FLAG antibody. We found that both WT and S129D ␣-syn failed to increase the ubiquitinated aggregates or form ubiquitinated inclusions (Fig. 5, A-I). Similar results were observed in neuro2a cells cotransfected with K48R or K63R mutant FLAG-Ub along with V5-lacZ, V5-tagged WT ␣-syn, or S129D ␣-syn (Fig. 4, C and D). These results suggest that the contribution of S129D-phosphorylated ␣-syn to the Ub-conjugates and ubiquitinated inclusions depends on the Lys 63 site not the Lys 48 site on Ub.
Phosphorylation of ␣-Syn at Ser 129 Increases Self-modification by Ub-Abundant deposition of ␣-syn and Ub in LBs may indicate an important role for ␣-syn-Ub linkage in the pathogenesis of LBs (9). However, using immunoprecipitation of GFP-tagged proteins followed by Western blotting with anti-FLAG antibody, we did not observe ubiquitinated ␣-syn in the soluble samples used in Figs. 2A and 4A (data not shown). To further explore a possible role for ␣-syn-Ub linkage in LB biogenesis, we examined the ubiquitination of ␣-syn within cells cotransfected with WT FLAG-Ub along with EGFP-tagged WT ␣-syn or S129D ␣-syn under the treatment of MG132. Neuro2a cells coexpressing ␤-syn and Ub were used as a control. Forty-eight hours after transfection, the cells were subsequently treated with MG132 (10 M) for 12 h. After treatment, cells were harvested and a coimmunoprecipitation assay was performed using anti-GFP antibody. The immunoprecipitants were subjected to immunoblot analysis using anti-FLAG antibody. As shown in Fig. 6, low migrating smeared bands compatible with ubiquitinated ␣-syn were observed. Smeared bands compatible with ubiquitinated ␤-syn were not detected (Fig.  6). Based on the apparent molecular mass of ␣-syn-EGFP (Ϸ47 kDa) and FLAG-Ub (Ϸ18 kDa), the high molecular mass species accounted for as mono-, di-, or tri-FLAG-Ub modified ␣-syn appear predominantly at the phosphomimic mutant (indicated by arrows in Fig. 6). These results support our earlier findings in this study that ␣-syn is co-localized with Ub in some of the aggregates. Considering the large amount of Ub-conjugates formed by the expression of S129D ␣-syn (Fig. 2, A and B), these results also suggest that phosphorylation of ␣-syn at Ser 129 may facilitate the ubiquitination of other proteins as well as itself.
Expression of S129D ␣-Syn Increases ␣-Syn Aggregate Formation under Oxidative Stress and Serum Deprivation-Lewy neurites, also known as dystrophic neurites, are one of the pathological characteristics of synucleinopathies. We therefore cultured neuro2a cells expressing EGFP, EGFP-tagged WT, or S129D ␣-syn in serum-free DMEM 24 h after transfection. Seventy-two hours later, the cells were observed using an inverted system microscope (Olympus, IX71, Japan). We observed that some of the surviving intact cells developed long neurites that are similar with the dystrophic neurites in morphous. It has been proposed that oxidative stress is associated with the pathogenesis of PD. A previous study suggests that oxidative stress induces the formation of ␣-syn inclusions (29). To explore whether phosphorylation of ␣-syn affects self-aggregation under oxidative stress, neuro2a cells were transfected with EGFP, EGFPtagged WT, or S129D ␣-syn for 24 h, followed by deprivation of serum and treatment with or without exposure to oxidative stress. Seventy-two hours later, the cells were observed using an inverted system microscope. The surviving cells with intact nuclei were selected for data analysis. We found that neuro2a cells overexpressing EGFP, EGFP-tagged WT, or S129D ␣-syn formed a few aggregates in the absence of serum (Fig. 7, A-D) with no statistic significance among the cells by quantitative data (Fig. 7J). However, oxidative stress induced the cytoplasmic and neuritic aggregates of ␣-syn in surviving neuro2a cells (Fig. 7, E-I) with cells overexpressing S129D ␣-syn forming approximately two times more ␣-syn aggregates than cells overexpressing WT ␣-syn (Fig. 7J). These results suggest that Ser 129 -phosphorylated ␣-syn may be more prone to aggregation under oxidative stress.

DISCUSSION
Typically, LBs are found in dopaminergic neurons in the substantia nigra pars compacta, but they are also observed in noradrenergic neurons in other brain regions and the peripheral nervous system (30,31). LBs have also been reported in many other neurological disorders, suggesting a common origination of LBs in various neurodegenerative diseases (3,4). Classical LBs are usually found with ␣-syn concentrated in the halo and Ub concentrated in the core (32). However, the exact mechanism that links ubiquitination to the formation of LB structure is poorly understood. In our cellular model, we demonstrate that coexpression of S129D ␣-syn with Ub in neuro2a cells forms ubiquitinated aggregates or inclusions. Their morphologies vary from small spherical Ub-positive punctates accompanied by diffusive or cloud-like ␣-syn distribution, to amorphous inclusion with Ub immunoreactivity and ␣-syn fluorescence, and to well defined co-localization of ␣-syn and Ub. These fluorescent spectra of ␣-syn and Ub  MAY 11, 2007 • VOLUME 282 • NUMBER 19

Contribution of Phosphorylated ␣-Syn to Accumulation of Proteins
staining are similar to the pathological variations observed in LBs from patients (32).
In our present study, we found that coexpression of WT or S129D ␣-syn with Ub increases the Lys 63 -linked Ub-conjugates and ubiquitinated aggregates in cultured cells. These results are consistent with the report that wild type and three familial PDrelated mutant ␣-syn lead to Lys 63 -linked Ub chain assembly in vitro (27). Because overexpression of WT ␣-syn in SHSY5Y cells has a considerable level of Ser 129 phosphorylation (29), it is possible that the assembly of Lys 63 -linked Ub-conjugates by WT ␣-syn results from internal phosphorylation at Ser 129 that may be mediated by G protein-coupled receptor kinases (33), casein kinase 1, and casein kinase 2 in vivo (34). Proteasomal inhibition resulting from ␣-syn plays an important role in the etiology of PD (35,36). Because a poly-Ub signal is believed to inhibit the proteasome (37,38), the increase in poly-Ub-conjugates by expression of S129D ␣-syn may suggest proteasomal inhibition by S129D ␣-syn, in turn leading to more and more poly-Ub-conjugates (39). Compared with WT ␣-syn, overexpression of S129D ␣-syn resulted in more poly-Ub-conjugates in the insoluble fraction. Because relevant studies in patients are mainly performed in Sarkosyl-insoluble fractions or extracts of LBs isolated from synucleinopathic brains (15,18), the amount of Ub-conjugates in the insoluble fraction may be more relevant to the LB-like structure formation in our cellular model. Taken together, these results indicate that the assembly of ubiquitinated proteins by phosphorylated ␣-syn may be important for LB biogenesis. A transgenic mouse model expressing S129D ␣-syn may be useful to further investigate the possible role of Ser 129 -phosphorylated ␣-syn in the pathogenesis of PD and other related diseases.
Ser 129 -phosphorylated ␣-syn, the dominant pathological ␣-syn species, is mono-, di-, or tri-ubiquitinated in brain extracts from synucleinopathic patients (15,18). The mono-or di-ubiquitinated ␣-syn species in the A53T ␣-Syn transgenic mice model further reveal a possible role for ubiquitinated ␣-syn in the neurodegeneration of synucleinopathy (40). Ubiquitination of ␣-syn has been demonstrated in vitro (41,42) as well as in cultured cells (41,43). After exposure to MG132, the modification of ␣-syn by mono-, di-, or tri-FLAG-Ub predominantly occurred at the phospho-mutant ␣-syn when probed by anti-FLAG antibody. This event suggests close ties between the ubiquitination of phospho-mutant ␣-syn to the ubiquitination of phosphorylated ␣-syn in human synucleinopathies, supporting the finding that phosphorylated ␣-syn in LBs is ubiquitinated (15,18). We speculate that the other smear bands shown in Fig. 6 may be due to other ubiquitinated proteins associated with ␣-syn. It seems that only a very small part of ␣-syn is ubiquitinated in the insoluble fraction in our cellular system or in a previously reported in vitro system (27). Therefore, we were unable to detect it using anti-GFP or anti-␣-syn antibody. These results may suggest that without inhibition of the proteasome, ubiquitinated ␣-syn levels may be too low to be detected in a cellular model in vitro. However, we cannot exclude the fact that the antibodies used in this study do not detect ubiquitinated ␣-syn species well.
Our data show that well defined colocalization of S129D ␣-syn and FLAG-Ub is only found in some cells with condensed ubiquitinated inclusions. As only a small part of ␣-synuclein is ubiquitinated in our cellular model as well as the in vitro assay reported by other investigators, the effect of phospho-␣synuclein on the accumulation of other ubiquitinated proteins, rather than itself ubiquitination, may play an important role in LB biogenesis.
tion of synphilin-1 depends on a high level of synphilin-1 (44). Because degradation of synphilin-1 is performed by UPS (42), it is possible that dysfunction of UPS caused by phosphorylated ␣-syn may increase the level of synphilin-1 and Lys 63 -linked ubiquitination of synphilin-1. This notion is supported by a recent report that the direct interaction between ␣-syn and parkin does not exist in vitro, but overexpression of WT ␣-syn restores the parkin-mediated synphilin-1 ubiquitination that is suppressed by 14-3-3⑀ in vivo (46). Notably, ␣-syn is involved in parkinmediated ubiquitinated inclusion formation (44). Ubiquitination of synphilin-1 mediated by another E3 ligase, SIAH, leads to cytosolic ubiquitinated inclusions in the presence of the proteasomal inhibitor (42). Thus, it is possible that dysfunction of UPS caused by ␣-syn may enhance the ubiquitination of synphilin-1 and other LB-associated proteins.
␣-Syn is prone to form fibrils in vitro (47,48). Overexpression of ␣-syn in cell lines and transgenic animal models proved that aggregation of ␣-syn plays a prominent role in the pathogenesis of synucleinopathies (49 -51). Therefore, it is generally accepted that LBs result from the aggregation of ␣-syn (52,53). Phosphorylation of ␣-syn at Ser 129 is believed to promote fibril formation in vitro (13). We found here that overexpression of S129D ␣-syn forms more ␣-syn aggregates induced by oxidative stress and serum deprivation than overexpression of WT ␣-syn does. This is consistent with the finding that oxidative stress induces phosphorylation of WT ␣-syn and ubiquitinated inclusion formation (29). Thus, it appears that the enhanced aggregation of S129D ␣-syn and not WT ␣-syn makes it possible for S129D ␣-syn to be abundant in ubiquitinated inclusions. We detected few or no ␣-syn aggregates under normal conditions, whereas S129D ␣-syn formed aggregates under oxidative stress. These observations suggest that oxidative stress may not only induce the phosphorylation of ␣-syn, but also provide the necessary factors required for ␣-syn aggregate formation.
Together, our results examine a hypothesis for the contribution of S129D ␣-syn to the biogenesis of LBs. Thus, phosphorylation and ubiquitination, two seemingly independent post-translation modifications of proteins, are involved in the convergent pathological scenario as a result of biological responses to environmental factors. Our findings here contribute toward understanding the role of FIGURE 7. Expression of S129D ␣-syn increases the aggregates of itself under oxidative stress. A-D, neuro2a cells expressing EGFP, EGFP-tagged WT ␣-syn, or S129D ␣-syn were cultured in serum-free DMEM 24 h after transfection. The cells were washed with PBS, fixed in PBS with 4% paraformaldehyde for 5 min, and treated with 0.25% Triton X-100 for 10 min. The cells were washed with PBS and incubated with DAPI for 3 min at room temperature. Finally, the cells were washed with PBS and observed using an inverted system microscope. E-I, neuro2a cells were transfected with EGFP, EGFP-tagged WT ␣-syn, or S129D ␣-syn. Aggregation induction in transfected cells was performed. Then, the cells were observed using an inverted system microscope. The cytoplasmic and neuritic aggregates of S129D ␣-syn are shown by arrows. D represents a higher magnification of the rectangle in C; H represents the higher magnification of the rectangle in G. Bar, 20 M. J, the number of cells with aggregates relative to the cells expressing EGFP, EGFP-tagged WT ␣-syn, or S129D ␣-syn was analyzed by one-way analysis of variance test. Data are shown as mean Ϯ S.E. **, p Ͻ 0.01 versus EGFP; ##, p Ͻ 0.01 versus WT. For each sample, five to six random fields were selected for counting. OS, oxidative stress.