Effects of Oxidative and Nitrative Challenges on (cid:1) -Synuclein Fibrillogenesis Involve Distinct Mechanisms of Protein Modifications*

Filamentous inclusions of (cid:1) -synuclein protein are hallmarks of neurodegenerative diseases collectively known as synucleinopathies. Previous studies have shown that exposure to oxidative and nitrative species stabilizes (cid:1) -synuclein filaments in vitro , and this stabilization may be due to dityrosine cross-linking. To test this hypothesis, we mutated tyrosine residues to pheny-lalanine and generated recombinant wild type and mutant (cid:1) -synuclein proteins. (cid:1) -Synuclein proteins lacking some or all tyrosine residues form fibrils to the same extent as the wild type protein. Tyrosine residues are not required for protein cross-linking or filament stabilization resulting from transition metal-mediated oxidation, because higher M r SDS-resistant oligomers and fil- aments stable to chaotropic agents are detected using all Tyr 3 Phe (cid:1) -synuclein mutants. By contrast, cross-linking resulting from exposure to nitrating agents required the presence of one or more tyrosine residues. Furthermore, tyrosine cross-linking is involved in filament stabilization, because nitrating agent-exposed assembled wild type, but not mutant (cid:1) -synuclein lacking all tyrosine residues,

Tyr residues in ␣-syn are targets for nitration (25), we investigated the role of each of the four Tyr residues (at amino acid positions 39, 125, 133, and 136) in ␣-syn fibrillization and fibril stabilization under oxidative and/or nitrative conditions. Using WT and Tyr (Tyr 3 Phe) mutant ␣-syn recombinant proteins in vitro and stably transfected cultured cells, we showed that the presence of one or more Tyr residues is required for nitrating species-induced ␣-syn protein cross-linking and filament stabilization. Surprisingly, ␣-syn protein cross-linking and fibril stabilization induced by transition metal-mediated oxidation does not require Tyr residues in ␣-syn. These data suggest that independent pathogenic mechanisms are involved in oxidativeversus nitrative-induced ␣-syn fibrillogenesis.
Exposure of ␣-Syn Proteins to the Nitrating Agent, Peroxynitrite-Peroxynitrite (ONOO Ϫ ) was synthesized from sodium nitrite and acidified H 2 O 2 , and excess H 2 O 2 was removed by treatment with manganese dioxide (31). Treatment of ␣-syn proteins with ONOO Ϫ was carried out as described previously (25). Briefly, ␣-syn proteins were diluted into nitration buffer containing 100 mM potassium phosphate, 25 mM sodium bicarbonate, pH 7.4, and 0.1 mM diethylenetriamine pentaacetic acid. The concentration of ONOO Ϫ was determined spectrophotometrically at 302 nm in 2 N NaOH ( 302 ϭ 1670 M Ϫ1 cm Ϫ1 ). Peroxynitrite was added in 2 boluses of 10-fold molar excess of the protein. After treatment, proteins were boiled in SDS sample buffer (10 mM Tris, pH 6.8, 1 mM EDTA, 40 mM dithiothreitol, 1% SDS, 10% sucrose) and resolved by SDS-PAGE. Proteins were either stained with Coomassie Blue R-250 for quantification by densitometry or analyzed by immunoblotting using monoclonal antibodies directed against WT ␣-syn and nitrated ␣-syn followed by incubation with an anti-mouse horseradish peroxidase-conjugated antibody (Jackson ImmunoResearch Laboratories, West Grove, PA) and detected with enhanced chemiluminescence reagents (PerkinElmer Life Sciences).
Oxidation of ␣-Syn Proteins Using Transition Metals and Peroxide-The effects of oxidative damage on ␣-syn proteins were examined in reaction mixtures of phosphate-buffered saline (137.5 mM NaCl, 2.5 mM KCl, 10 mM Na 2 HPO 4 , 1.75 mM KH 2 PO 4 , pH 7.5) in the presence of 500 or 50 M CuCl, CuCl 2 , FeCl 2 , or FeCl 3 and 300 M H 2 O 2 . Where indicated, reactions were incubated at 37°C for 4 h and prepared for assembly experiments as described below or diluted with SDS sample buffer and heated to 100°C for 5 min.
Assembly and Polymerization of ␣-Syn Proteins-Untreated or ONOO Ϫ -exposed ␣-syn proteins at a concentration of 5 mg/ml were exchanged in 100 mM sodium acetate buffer (pH 7.0) containing 0.04% sodium azide. Studies involving copper and peroxide oxidation were similarly exchanged into acetate buffer, and the oxidative reagents were added at a final concentration of 500 M CuCl and 300 M H 2 O 2 . ␣-Syn proteins were incubated either at 33°C for 4 days or at 37°C for 2 days with continuous shaking at 1000 rpm. Each assembly reaction sample was overlaid with 40 -50 l of mineral oil to prevent condensation of samples and hence the alteration of results.
Direct Visualization of ␣-Syn Fibril Formation-For the direct ultrastructural inspection of filament formation, ␣-syn protein samples were adsorbed to 300-mesh carbon-coated copper grids, stained with 1% uranyl acetate, and visualized with a transmission electron microscope (Joel 1010, Peabody, MA) as described previously (19).
Assessment of ␣-Syn Polymerization by Centrifugal Sedimentation Analysis-Samples were centrifuged at 100,000 ϫ g for 20 min after assembly incubation. Supernatants and pellets were separated, SDS sample buffer was added, and the samples were heated to 100°C for 5 min. ␣-Syn proteins were resolved by SDS-PAGE, stained with Coomassie Blue R-250, and quantified by densitometry. The kinetic consistency of the experimental conditions used for fibril formation was determined by incubating ␣-syn proteins for varying lengths of time (3 to 96 h) at 37°C and 1000 rpm using sedimentation analysis in addition to the K114 fluorescence analysis (described below) to confirm that ␣-syn protein assembly increases over time reproducibly.
Assessment of the Formation of Amyloidogenic Polymers Using the K114 Assay-The formation of ␣-syn amyloidogenic polymers was assessed using a novel amyloid-binding dye (trans,trans)-1-bromo-2,5-bis-(4-hydroxy)styrylbenzene (or K114). K114 recognizes amyloid fibrils in vitro and in pathological inclusions and has distinct fluorescent properties that allow for the monitoring of amyloid fibril formation in solution. 2 The K114 fluorescence assay was performed by adding 5 l of each assembly sample (25 g of ␣-syn protein) to 100 l of K114 assay solution (100 mM glycine buffer, pH 8.5, 50 M K114). Fluorescence was measured using a SpectraMax GeminiXS fluorometer (Molecular Devices, Sunnyvale, CA) and SoftMax Pro 4.0 software with a fixed excitation wavelength of 380 nm and a fixed emission wavelength of 550 nm with a cutoff at 530 nm.
Filament Stabilization Studies-Following ␣-syn protein assembly at 37°C for 3 days with continuous shaking, samples were centrifuged at 100,000 ϫ g for 20 min. Pellets were resuspended in nitration buffer and exposed to ONOO Ϫ at room temperature or resuspended in phosphate-buffered saline and treated with 500 M CuCl and 300 M H 2 O 2 for 1 h at 37°C. Samples were then diluted into either H 2 O or 4 M urea, incubated at room temperature for 10 min, and centrifuged at 100,000 ϫ g for 20 min. The supernatants and pellets were separated, and samples were heated to 100°C for 5 min in SDS sample buffer. ␣-Syn proteins were resolved by SDS-PAGE, stained with Coomassie Blue R-250, and quantified by densitometry.
In Vivo Oxidative/Nitrative Stress Experiments-Human ␣-syn cDNAs corresponding to the WT or the mutant protein with all four Tyr residues mutated to Phe, i.e. Y39F,Y125F,Y133F,Y136F (referred to as 4(Y 3 F)), were cloned into the pcDNA 3.1ϩ mammalian expression vector (Invitrogen). HEK293 cells were maintained with Dulbecco's modified medium-high glucose containing 10% fetal bovine serum, 1% penicillin/streptomycin, and 2% L-glutamine. Cells were transfected with WT and 4(Y 3 F) ␣-syn plasmids by a calcium phosphate precipitation method buffered with BES (32). Stably transfected clones were selected for and maintained in culture medium supplemented with 300 g/ml geneticin. Cells were plated at a density of 1 ϫ 10 6 /well in 35-mm tissue culture plates and exposed to PapaNO (1-propananmine-3-(2hydroxy-2-nitroso-1-propylhydrazine)) and/or paraquat as described previously (33). Briefly, cells were rinsed with warm media and then incubated in 2 ml of media supplemented with 7.5 mM paraquat for 30 min at 37°C. Subsequently, PapaNO was added at a final concentration of 1.05 mM, and cells were incubated for an additional 1.5 h.
Cells plated on coverslips were fixed in 4% paraformaldehyde and permeabilized in 0.2% Triton/phosphate-buffered saline. Cells on coverslips were incubated overnight at 4°C with Syn 208 and Syn 204, rinsed the following morning, and incubated with secondary anti-mouse antibody Alexa 488 fluorescein isothiocyanate (Molecular Probes, Eugene, OR). Coverslips were mounted on slides using Vecta Shield containing 4,6-diamindino-2-phenylindole (Vector Laboratories, Burlingame, CA). Cells plated without coverslips were rinsed with phosphate-buffered saline and lysed in 2% SDS and 50 mM Tris, pH 6.8. Protein concentration was determined using the bicinchoninic acid protein assay kit. Samples were boiled with SDS sample buffer, and 7 g of total protein were resolved by SDS-PAGE and immunoblotted with LB509.

RESULTS
Role of Tyr Residues in the Fibrillogenesis of ␣-Syn Proteins-To evaluate the potential role of Tyr residues in ␣-syn fibril formation, recombinant human ␣-syn proteins with one to all four of the Tyr residues mutated to Phe were expressed in E. coli and purified (Fig. 1A). The ability of these WT and mutant FIG. 1. Tyr residues are not required for ␣-syn filament assembly. A, Coomassie Blue R-250-stained SDS-polyacrylamide gel of purified human recombinant WT and Tyr 3 Phe ␣-syn proteins. Five hundred ng of each protein was loaded on the gel. B, ␣-syn proteins at a concentration of 5 mg/ml were assembled in 100 mM sodium acetate, pH 7.0, at 33°C for 4 days by continuous shaking at 1000 rpm and assessed by sedimentation analysis. S and P represent supernatant and pellet, respectively. C, quantitative sedimentation analysis of ␣-syn oligomerization as determined by densitometry of Coomassie-stained gels (n ϭ 5). D, the assembly of WT and 4(Y 3 F) ␣-syn at 1000 rpm and 37°C was monitored over time by K114 fluorescence and ultracentrifugation sedimentation analyses (n ϭ 5). E, assessment of ␣-syn proteins for filament formation by EM. Scale bar equals 500 nm. ␣-syn proteins to form fibrils was evaluated by several methods. Centrifugal sedimentation analysis was used as a quantitative method to monitor the formation of pelletable ␣-syn polymers, and this assay has previously been shown to parallel the formation of fibrils (19,34). However, filament formation can be more directly determined by negative staining electron microscopy (EM) (Fig. 1E). Furthermore, the K114 fluorescent assay was used to monitor the formation of amyloidogenic polymers. K114 is a congener of Congo Red with unique fluorescent properties and solubility that allow for the quantitative measurement of amyloid fibril formation in solution. 2 The formation of amyloid fibrils as measured by K114 fluorescence parallels the formation of sedimentable polymers as well as fibrils observed directly by transmission EM. 2 Therefore, these three assays compliment each other in assessing and confirming the formation of pelletable amyloidogenic ␣-syn fibrils.
All Tyr 3 Phe mutant proteins were capable of forming polymers to the same extent as the WT ␣-syn protein as assessed by sedimentation analysis (Fig. 1, B and C). Kinetic analysis of WT and mutant ␣-syn assembly using both sedimentation and K114 fluorescence analyses revealed that amyloidogenic polymer formation increased consistently and reproducibly over time, although the 4(Y 3 F) mutant ␣-syn protein was slower in forming amyloidogenic fibrils than the WT ␣-syn protein (Fig. 1D). Shaking at slightly higher temperatures results in a higher percentage (80 -90%) of the ␣-syn monomers converted to fibrils in both WT and mutant proteins (data not shown). Furthermore, direct observation of fibrils by EM examination failed to reveal any significant morphological difference between WT and mutant ␣-syn fibrils, including width, length, or twisting and turning of the fibrils (Fig. 1E). These results indicate that the hydroxyl functional group on Tyr residues in ␣-syn proteins is not critical for the assembly of the protein into fibrils under these in vitro conditions.

Role of Tyr Residues in the Cross-linking of ␣-Syn
Proteins by Nitrating Agents-Previous data indicated that WT ␣-syn protein exposed to nitrating agents forms oligomers through dityrosine cross-linking (25). However, the specific Tyr residues involved in the formation of dityrosine cross-linked oligomers have not been identified. To address this question, WT and mutant ␣-syn proteins were exposed to the nitrating agent, ONOO Ϫ , in the presence of CO 2 . The exposed ␣-syn proteins were resolved by SDS-PAGE gels and analyzed by Coomassie Blue R-250 staining ( Fig. 2A) or Western blotting using antibodies specific for the different nitrated Tyr residues (Fig. 2B). Exposure of WT, single, double, and triple Tyr 3 Phe mutant ␣-syn proteins to ONOO Ϫ in the presence of CO 2 induces nitration as well as oxidation of Tyr residues to form dityrosine, which results in the formation of SDS-stable dimers and oligomers (Fig. 2, A and B). These results suggest that all four Tyr residues are potential targets for protein nitration as well as for dityrosine cross-linking. As predicted, the mutant ␣-syn FIG. 2. ONOO ؊ nitrating agent induces cross-linking but blocks ␣-syn filament formation. A, ␣-syn proteins were exposed to ONOO Ϫ , resolved on 12% polyacrylamide gels, and visualized by staining with Coomassie Blue R-250. B, proteins exposed to ONOO Ϫ were transferred to nitrocellulose membranes and analyzed by Western blotting using monoclonal antibodies that recognize native ␣-syn (Syn 208) and nitrated ␣-syn (nSyn 12, nSyn 14, and nSyn 24). The specificities of the anti-nitrated ␣-syn antibodies are described under "Experimental Procedures." Using these anti-nitrated site-specific ␣-syn antibodies, we demonstrate the specificity of the mutant ␣-syn proteins used here as well as the site(s) that are nitrated. C, sedimentation analysis of the effect of ONOO Ϫ exposure on the ability of ␣-syn to assemble into fibrils. Proteins treated with ONOO Ϫ were incubated under assembly conditions and sedimented at 100,000 ϫ g for 20 min. Samples were resolved on 12% SDS-polyacrylamide gels and stained with Coomassie Blue R-250. S and P represent supernatants and pellets, respectively. The bar graph in D shows percentage of WT and 4(Y 3 F) ␣-syn proteins found in pellet after sedimentation analysis of samples in C (n ϭ 6).
protein that lacks all four Tyr residues (4(Y 3 F)) neither formed SDS-resistant dimers nor was nitrated after exposure to the nitrating agent (Fig. 2, A and B). Decreasing the number of Tyr residues in the ␣-syn protein resulted in a reduction in SDS-stable oligomer, yet the presence of a single Tyr residue in the ␣-syn protein (Y125F,Y133F,Y136F; Y39F,Y133F,Y136F; Y39F,Y125F,Y133F; and Y39F,Y125F,Y136F) is sufficient for the formation of dimers (Fig. 2B).
The ability of WT and mutant ␣-syn proteins to polymerize after exposure to ONOO Ϫ in the presence of CO 2 was determined by sedimentation analysis (Fig. 2, C and D). The formation of pelletable polymers of WT and mutant ␣-syn proteins containing at least one intact Tyr residue was inhibited by treatment with ONOO Ϫ in the presence of CO 2 . The lack of ␣-syn fibrils from samples of ONOO Ϫ -exposed WT and mutant ␣-syn proteins containing at least one intact Tyr residue was further confirmed by EM (data not shown), which indicates that the nitration of any Tyr residue can contribute to the prevention of fibril formation. The 4(Y 3 F) ␣-syn protein exposed to ONOO Ϫ formed some pelletable oligomers (Fig. 2, C and D), but to a lesser extent than unexposed 4(Y 3 F) ␣-syn. This result indicates that other protein modifications, in addition to 3-nitrotyrosine, contribute to the inhibition of the polymerization of ␣-syn protein upon exposure to nitrating agents. Indeed, nitrating agents such as peroxynitrite or nitrogen dioxide are higher oxidation-state species of nitric oxide, formed by chemical reactions of nitric oxide with superoxide or by the one electron oxidation of nitrite by peroxidases and other heme proteins and hydrogen peroxide. Although these reactive nitrogen species primarily nitrate Tyr residues, they are also capable of oxidizing Tyr to form dityrosine cross-linked species. In addition, other amino acid residues, such as cysteine, tryptophan, and histidine could be oxidized. Therefore, exposure of ␣-syn protein to nitrating agents results in both nitration and some oxidation (25).
Tyr Residues of ␣-Syn Proteins Are Not Critical for the Oligomerization Induced by Oxidizing Agents-It was previously shown that WT recombinant ␣-syn protein forms SDS-stable cross-linked oligomers upon exposure to oxidizing agents such as H 2 O 2 and redox active metals (26,35,36). Consistent with these previous reports, incubation of WT ␣-syn with CuCl plus H 2 O 2 resulted in the formation of SDS-stable oligomers (Fig.  3A), whereas incubation with transition metals or H 2 O 2 separately did not result in the formation of these oligomeric species (data not shown). In contrast to the effects of nitration, all mutant ␣-syn proteins formed cross-linked oligomers equally well, including the 4(Y 3 F) mutant ␣-syn protein (Fig. 3A). Replacing CuCl with FeCl 2 , FeCl 3 , or CuCl 2 in the presence of H 2 O 2 resulted in similar formation of protein oligomers (data not shown). Furthermore, WT and 4(Y 3 F) ␣-syn proteins exposed to CuCl and H 2 O 2 were also able to assemble and form fibrils after shaking at 1000 rpm for 2 days at 37°C, and a time course study was performed using centrifugal sedimentation analysis in conjunction with the K114 fluorescence technique to ensure the reproducibility of our results (Fig. 3B). WT and 4(Y 3 F) ␣-syn proteins assembled to the same extent with or without CuCl and H 2 O 2 treatment, irrespective of whether or not the protein was oxidized during the assembly (Fig. 3, C and D) or 4 h before the assembly (data not shown). These data indicate that simple oxidation without nitration of the protein does not target Tyr residues, and it does not render ␣-syn assembly incompetent. EM analysis was conducted to determine the precise structural characteristics of these oxidationchallenged ␣-syn samples. The micrographs depict the samples as bundles of tightly associated filaments admixed with some non-filamentous ␣-syn protein (Fig. 3E). The formation of both filamentous and non-filamentous aggregates is likely because of the co-occurrence of filament assembly and chemical crosslinking of non-fibrillar ␣-syn because of oxidation. Taken together, these results suggest that Tyr nitration disrupts the process of fibril formation, but oxidation without nitration of Tyr residues does not prevent ␣-syn protein fibril formation and may potentiate the formation of inclusions that are comprised of both fibrillar and aggregated ␣-syn protein.
Role of Tyr Residues in the Stability of WT and 4(Y 3 F) ␣-Syn Proteins-The ability of nitrating and oxidizing agents to stabilize the pre-assembled, filamentous WT and 4(Y 3 F) ␣-syn proteins was evaluated by incubating pre-assembled proteins exposed to nitrating and/or oxidizing agents with the chaotropic agent, 4 M urea. The addition of 4 M urea to fibrillar ␣-syn disrupts the majority of the fibrils so that most of the protein is recovered from the 100,000 ϫ g soluble fraction (Fig.  4A). EM confirmed the disruption of ␣-syn fibrils such that after exposure to 4 M urea, short fibrils were only very rarely observed (Fig. 4F). Consistent with our previously report (25), exposure of pre-assembled WT ␣-syn protein to ONOO Ϫ resulted in the formation of urea-stable polymers (Fig. 4A). Peroxynitrite-exposed assembled WT ␣-syn protein is significantly more stable to treatment with chaotropic agents compared with untreated WT ␣-syn protein, where the urea-stable pelletable fraction increases from 23.6 to 97.5% upon treatment with ONOO Ϫ (Fig. 4, A and E). Consistently, EM assessment of samples of WT ␣-syn fibrils exposed to ONOO Ϫ followed by 4 M urea revealed that, although the amount of fibrils was slightly reduced compared with untreated/unsolubilized sample, there were still abundant fibrils that were resistant to urea (Fig. 4F). By contrast, exposure of pre-assembled 4(Y 3 F) mutant ␣-syn fibrils to ONOO Ϫ failed to render them resilient to 4 M urea by sedimentation analysis (Fig. 4, B and E), and there were far fewer urea-resistant fibrils observed by EM (data not shown). These data are consistent with the hypothesis that dityrosine cross-linking is responsible for the stabilization of ␣-syn fibrils treated with chaotropic agents. Thus, Tyr residues are primary sites of reactivity for nitrating agents with ␣-syn protein. By contrast, exposure of both assembled WT and 4(Y 3 F) ␣-syn proteins to oxidants (Cu/H 2 O 2 ) caused an increase in the ureastable pelletable fraction from 11.2 to 73.55% for WT protein and 32.4 to 76.75% for 4(Y 3 F) ␣-syn protein (Fig. 4, C-E). Thus, oxidation of assembled ␣-syn proteins induced fibril stability to disruption by chaotropic agents through a Tyr-independent oligomerization mechanism.

Formation of WT, but Not 4(Y 3 F), ␣-Syn Protein Inclusions by Intracellular Nitrative Insults-Our previous studies
have demonstrated that treatment of HEK293 cells stably expressing WT ␣-syn protein with nitrating and oxidizing agents resulted in the formation of cytoplasmic nitrated ␣-syn inclusions (33). To assess the role of Tyr residues in this process, we treated HEK293 cells stably expressing either WT or 4(Y 3 F) ␣-syn protein with paraquat, an alkylating agent and superoxide generator (37)(38)(39), and PapaNO, a nitric oxide donor (40,41). The formation of small ␣-syn inclusions in the cytoplasm of cells was assessed by indirect immunofluorescence (Fig. 5A). As expected, cells expressing WT ␣-syn protein formed ␣-syn protein inclusions in ϳ3-4% of the cells, whereas cells expressing 4(Y 3 F) ␣-syn protein produced significantly less of these ␣-syn inclusions (ϳ0.2%) (Fig. 5B). This finding was not a result of protein expression levels as the 4(Y 3 F) ␣-syn cells express more ␣-syn protein than the WT ␣-syn cells (Fig. 5C). Moreover, the ␣-syn protein levels were not influenced by treatment with paraquat and nitric oxide reagents (Fig. 5C). Untreated cells and cells treated with paraquat alone or PapaNO alone did not show ␣-syn inclusion formation (Fig. 5B). Although cells treated under these nitrative and oxidative conditions were lysed and fractionated to determine the formation of covalently linked oligomers, higher M r oligomers were not detected (Fig. 5D), even after prolonged exposure of Western blots (data not shown). These results indicate that if cross-linking occurs, it must be at very low levels. Nevertheless, the paucity of inclusions in cells expressing 4(Y 3 F) ␣-syn suggests that the nitrative modification of Tyr residues in ␣-syn induces a redistribution of the protein into focal accumulations. DISCUSSION Since the identification of ␣-syn protein as the building block for LBs and GCIs in neurodegenerative diseases, the mechanism whereby this abundant protein converts from a highly soluble form to insoluble filamentous pathological inclusions has been the subject of intense investigation (13)(14)(15). ␣-Syn readily forms fibrils in vitro, and the process of fibril formation is modulated by factors such as pH, temperature, and divalent metals (19, 24, 26, 34 -36, 42, 43). Biochemical analysis of ␣-syn extracted from the brains of patients with these disorders re-

FIG. 3. Oxidation of ␣-syn proteins by CuCl and H 2 O 2 induces cross-linking and oligomerization independent of Tyr residues.
A, ␣-syn proteins were exposed to 300 M H 2 O 2 and 500 M CuCl for 4 h at 37°C. Samples were resolved on 12% SDS-polyacrylamide gels and immunoblotted with Syn 208 antibody directed against ␣-syn to analyze the formation of SDS-stable oligomers. B, the simultaneous assembly and incubation of WT and 4(Y 3 F) ␣-syn with 300 M H 2 O 2 and 500 M CuCl at 1000 rpm and 37°C was monitored over time by K114 fluorescence and ultracentrifugation sedimentation analyses (n ϭ 5). C, WT and 4(Y 3 F) ␣-syn proteins were assembled alone (Ϫ) or in the presence of 300 M H 2 O 2 and 500 M CuCl (ϩ) for 2 days. Samples were centrifuged to separate supernatants (S) from pellets (P) and were visualized by Coomassie Blue R-250 staining after running on 15% SDS-polyacrylamide gels. Note the dimer formation in treated samples. The bar graph in D shows the percentage of ␣-syn proteins found in the pellet after centrifugation and densitometric analysis of samples in C (n ϭ 4). E, ultrastructural EM analysis of WT ␣-syn after concurrent incubation with 300 M H 2 O 2 and 500 M CuCl and assembly at 1000 rpm and 37°C. Note the presence of tight bundles of filamentous as well as non-filamentous proteinaceous aggregates. Scale bar equals 100 nm.
vealed the presence of ␣-syn oligomers that are stable to chaotropic agents as well as to SDS, suggesting that these ␣-syn oligomers are covalently cross-linked (8,11,23,24). The in vitro nitration/oxidation of Tyr residues to form dityrosine suggests one potential mechanism for the formation of covalently linked ␣-syn oligomers (25). Indeed, we demonstrated previously that ϳ30% of ␣-syn is cross-linked by dityrosine after exposure to nitrating agents such as ONOO Ϫ /CO 2 and myeloperoxidase/H 2 O 2 /nitrite (25). Similarly, Paik et al. (26,36) suggested that ␣-syn dityrosine cross-linking can be caused by H 2 O 2 and redox active metals such as copper, which in turn leads to oligomerization of the protein. The possibility that nitrative/oxidative mechanisms may be involved in the formation of pathological inclusions comprised of insoluble, filamen-tous, and potentially cross-linked ␣-syn is supported by the demonstration that ␣-syn nitrated at multiple Tyr residues is recovered from brains containing LBs, GCIs, and Lewy neurites, and monoclonal antibodies recognizing only nitrated ␣-syn protein decorates these protein inclusions in Parkinson's disease, dementia with LBs, multiple system atrophy, and neurodegeneration with brain iron accumulation type I (23). Collectively, these results support the idea that reactive species capable of nitrating and possibly cross-linking Tyr residues in ␣-syn may represent a pathogenic mechanism leading to synucleinopathies, and we propose that the nitrative alterations of Tyr residues in ␣-syn may play a critical role in the formation of ␣-syn filamentous polymers.
To test our hypothesis, we examined the critical role of Tyr   FIG. 4. Nitration-induced fibril stability is dependent on Tyr residues, but oxidation-mediated fibril stabilization is not. Filamentous ␣-syn proteins were assembled and isolated by incubation at 37°C for 3 days followed by sedimentation at 100,000 ϫ g for 20 min. Unassembled or filamentous WT and 4(Y 3 F) ␣-syn proteins were exposed to either 10 times molar excess ONOO Ϫ or 300 M H 2 O 2 and 500 M CuCl. Fibril stability was assessed by the addition of 4 M urea for 10 min, followed by sedimentation at 100,000 ϫ g for 20 min. Supernatants (S) and pellets (P) were resolved on 15% SDS-polyacrylamide gels and analyzed by Western blotting with LB509. Unassembled or assembled WT (A and C) and 4(Y 3 F) (B and D) ␣-syn proteins were untreated (Ϫ) or treated (ϩ) with ONOO Ϫ at room temperature (A and B) or 300 M H 2 O 2 and 500 M CuCl at 37°C for 1 h (C and D). The bar graph in E shows percentages of ␣-syn proteins found in pellets after centrifugal sedimentation analysis of ONOO Ϫ -or CuCl/H 2 O 2treated and untreated assembled WT and 4(Y 3 F) ␣-syn proteins were incubated with 4 M urea (n ϭ 2). The electron micrographs in F depict untreated assembled WT ␣-syn protein incubated with 4 M urea and ONOO Ϫ -exposed assembled WT ␣-syn protein incubated with 4 M urea. Scale bar equals 100 nm. residues in ␣-syn fibril formation and stabilization after exposure to nitrating and/or oxidizing agents. Our results showed that Tyr residues are responsible for the formation of covalently cross-linked ␣-syn dimers and oligomers after exposure to ONOO Ϫ , because such dimers were not detected in the 4(Y 3 F) ␣-syn mutant protein that lacks Tyr residues. By contrast, Tyr residues are not responsible for the formation of covalently linked dimers and higher order oligomers after treatment of ␣-syn with pure oxidizing agents (e.g. H 2 O 2 plus CuCl), because these stable oligomers were detected in the 4(Y 3 F) ␣-syn mutant protein. Furthermore, despite the formation of dityrosine dimers, ␣-syn proteins nitrated at one or more Tyr residues do not assemble into fibrils, whereas both native and purely oxidized ␣-syn proteins are capable of fibrillogenesis under the same experimental conditions. Because 4(Y 3 F) ␣-syn was able to fibrillize following exposure to nitrating agents, albeit to a lesser extent compared with untreated control, this suggests that nitration is at least partially responsible for the inhibition of fibril formation. It is important to recognize that 4(Y 3 F) ␣-syn protein was affected somewhat by the ONOO Ϫ exposure, as it did not fibrillize to control levels. This result may be accounted for by other oxidative mechanisms associated with ONOO Ϫ , and these alternative chemical modifications act synergistically FIG. 5. Tyr residues facilitate the formation of ␣-syn inclusions in HEK293 cells upon exposure to ONOO ؊ -generating agents. HEK293 cells stably overexpressing WT and 4(Y 3 F) ␣-syn were exposed to 7.5 mM paraquat and 1.05 mM PapaNO to generate ONOO Ϫ intracellularly. In A, HEK293 cells were fixed on coverslips and stained with monoclonal anti-␣-syn antibodies Syn 204 and Syn 208 (green). Nuclei are stained blue with 4,6-diamindino-2-phenylindole. Inclusions are indicated by arrows. In B, a bar graph shows percentage of HEK293 cells containing ␣-syn inclusions in the following conditions: untreated, paraquat alone, PapaNO alone, and paraquat and PapaNO together; * indicates statistical significance between WT ␣-syn-untreated cells and WT ␣-syn cells treated with both paraquat and PapaNO, p Ͻ 0.005; ** indicates statistical significance between 4(Y 3 F) ␣-syn cells treated with both paraquat and PapaNO and WT ␣-syn cells treated with paraquat and PapaNO, p Ͻ 0.01. In C and D, untreated (Ϫ) and paraquat-and PapaNO-treated (ϩ) WT and 4(Y 3 F) ␣-syn HEK293 cell lysates were analyzed by immunoblotting with LB509. Equal amounts of protein (7 g) were resolved on a 15% SDS-polyacrylamide gel. Equal protein loading was confirmed by monitoring the levels of ␣-tubulin in C. In D, ␣-syn oligomers are not detected by Western blot analysis in cell lysates of treated (ϩ) and untreated (Ϫ) WT and 4(Y 3 F) ␣-syn HEK293 cells.
in preventing the polymerization of ␣-syn. Although oxidation caused by CuCl and H 2 O 2 does not affect the assembly of ␣-syn proteins, the oxidation caused by ONOO Ϫ may in fact be of a different mechanism and hence affect fibril formation. In this regard, our observations with nitrated ␣-syn might be reminiscent of the recent report demonstrating that dopamine-mediated oxidative alterations of ␣-syn prevent the formation of mature fibrils, leading to the accumulation of structures consistent with protofibrils (27). Because it is difficult to identify and quantify protofibrils, we are not certain whether or not the nitration-induced inhibition of ␣-syn fibrillogenesis acts through a similar mechanism. Future studies will be needed to address this important issue.
As noted above, both WT and 4(Y 3 F) ␣-syn proteins were unaffected in their abilities to form cross-linked oligomers and fibrils when exposed solely to oxidizing agents, i.e. Cu/H 2 O 2 (Fig. 3), indicating that this oxidation of Tyr residues and/or other ␣-syn amino acid residues does not affect the ability of the protein to form fibrils. Uversky et al. (43) reported that extensive oxidation of all methionine residues in ␣-syn protein using 1.2 M H 2 O 2 prevents fibril formation of the protein, yet this oxidation process is likely to be of a different mechanism or the treatment with 1.2 M H 2 O 2 more extensively oxidizes the protein compared with the system used in the studies of this paper. Therefore, the degree of oxidation may also be critical for the ability of WT ␣-syn to form fibrils.
Recently, Krishnan et al. (44) reported that the formation of dityrosine dimers can facilitate fibrillogenesis, and they suggested that these dimers might be a critical step in this process. Our studies with 4(Y 3 F) ␣-syn protein demonstrate that fibrillogenesis does not require the formation of dityrosine, but this does not exclude that dityrosine dimers could potentially act as initiators of polymerization (44). A major difference between our studies and those of Krishnan et al. (44) is the condition used to promote fibrillogenesis. Our system involves constant agitation to increase the rate of polymerization, which allows our data to be measured and analyzed within hours. However, the authors of the aforementioned work used a non-agitation technique, which requires much longer periods of time (15-20 days) for ␣-syn to fibrillize (44). These differences in assembly kinetics could be responsible for the apparent differences in the requirement for dityrosine dimers to initiate filament formation. The most parsimonious interpretation of these data is that filament formation does not require dityrosine dimers, but the formation of these dimers can facilitate fibril formation, especially under slow kinetic conditions. Exposure of ␣-syn to the nitrating agent, ONOO Ϫ , stabilizes filaments assembled from WT but not 4(Y 3 F) ␣-syn proteins. In contrast, oxidative modifications induced by CuCl and H 2 O 2 stabilize fibrils comprised of either WT or 4(Y 3 F) ␣-syn proteins. This result suggests that different mechanisms must be responsible for the stabilization of ␣-syn fibrils by nitrative and oxidative damage. Whereas Tyr residues are required for the stabilization of ␣-syn filaments induced by ONOO Ϫ , the stabilization of ␣-syn fibrils induced by CuCl-and H 2 O 2 -mediated oxidative insults is independent of Tyr residues. Further investigation into the oxidative-induced modifications of ␣-syn will be needed to determine which amino acid residues are affected in this process.
The involvement of Tyr residues in the formation of ␣-syn inclusions in vivo associated with nitrative and oxidative stress was assessed in a cultured cell line. Previously we showed that exposure of HEK293 cells expressing WT as well as the A53T and A30P ␣-syn mutant proteins to nitric oxideand superoxide-generating compounds such as paraquat, rotenone, dopamine, and PapaNO resulted in the formation of small cytoplasmic ␣-syn inclusions (33). These cellular inclusions contain nitrated ␣-syn, and ultrastructural analysis showed evidence for the presence of ␣-syn filaments (33), suggesting that nitration may play a role in the stabilization of some small ␣-syn oligomers present in HEK293 cells ex-FIG. 6. Proposed model for the role of nitration and oxidation in the formation of stable ␣-syn fibrils. Soluble ␣-syn can be induced to form cross-linked oligomers by nitrative stress (1) or oxidative stress (2). However, nitration-induced ␣-syn cross-link oligomers and nitrated monomers are not capable of assembling into fibrils (3). ␣-Syn monomers and oligomers modified by metal-mediated oxidation are still capable of forming fibrils, and the initial step of this process requires a structural change from ␣-helix and/or random coil to ␤-pleated sheet conformation (4). ␣-Syn in this ␤-pleated sheet conformation can assemble into protofibrils and eventually into fibrils (5). Both oxidative and nitrative modifications resulting in covalent cross-links can stabilize ␣-syn filaments (6), which can aggregate with other cellular proteins to form proteinaceous inclusions (7) and hence form LBs, Lewy neurites, and/or GCIs found in many synucleinopathies. q, soluble ␣-syn in ␣-helical and/or random coil conformation; N, nitration; O, oxidation; f, ␣-syn in a ␤-pleated sheet conformation;ˆ, covalent cross-links; OE, छ, E, other cellular proteins found with ␣-syn protein in synucleinopathy inclusions.
pressing very high levels of ␣-syn protein. Because our transfected cells express high levels of ␣-syn protein and it is known that ␣-syn protein has a tendency to assemble into fibrils at high concentrations, a dynamic equilibrium may exist between the soluble monomers of ␣-syn and protofibrils and small oligomers of ␣-syn intracellularly. The equilibrium may favor monomeric ␣-syn in the absence of nitrating agents, but in the presence of nitrating agents, oligomers may become stabilized by nitration and/or dityrosine crosslinking and hence form a nidus for further ␣-syn inclusion formation. The demonstration that HEK293 cells stably expressing 4(Y 3 F) ␣-syn produce significantly fewer inclusions when challenged with nitric oxide and superoxide generators (Fig. 5, A and B) provides further support that nitration of Tyr residues in ␣-syn is involved in the stabilization of these inclusions. It is unclear whether or not dityrosine formation is involved, because we were unable to detect any cross-linked species of ␣-syn (Fig. 5D). However, because only a small number of cells (3-4%) developed ␣-syn aggregates and because it is likely that only a small amount of ␣-syn is modified by nitration and/or dityrosine crosslinking, it is not surprising that we were unable to detect dityrosine by Western blot analyses. Future development of a specific antibody to dityrosinated ␣-syn may help to resolve this issue.
Recently, Fujiwara et al. (45) reported the extensive phosphorylation of Ser-129 in urea-soluble monomeric ␣-syn protein extracted from LBs in human disease brain. However, these authors did not find any modification at Tyr residues 133 and 136 (the sequence data did not cover Tyr residues 39 or 125). With the use of specific anti-nitrated ␣-syn antibodies, we have shown that LBs are specifically labeled with antibodies recognizing Tyr-39 as well as Tyr residues 125, 133, and 136 (23). The discrepancy between our data and that of Fujiwara et al. (45) could be explained as follows. We recovered the majority of nitrated ␣-syn in the urea-insoluble fraction, but Fujiwara et al. (45) did not analyze this insoluble fraction. Also, because nitrating agents would also covalently cross-link ␣-syn, nitrated ␣-syn molecules would be present in both monomeric and cross-linked oligomeric protein, but this also was not examined in the aforementioned work.
Our studies demonstrate for the first time that distinct oxidative and nitrative mechanisms play a role in modulating ␣-syn fibril formation and stabilization. Fig. 6 illustrates the possible consequences of nitrative-and oxidative-induced modifications on the formation of ␣-syn lesions. Tyr residues are required for nitrating agent-induced ␣-syn covalent cross-linking and nitration. Although dityrosine cross-linking could represent a critical step in fibril formation, the nitration of Tyr residues appears to prevent fibrillogenesis from soluble ␣-syn proteins. Once fibrils of ␣-syn protein are formed, nitrating agents can stabilize these pre-formed filaments through dityrosine cross-linking. These data suggest that the nitrated filamentous ␣-syn protein detected in LBs, Lewy neurites, and GCIs is likely to be a late event that occurs after ␣-syn fibrils already have developed. On the other hand, oxidative stress could be an early or late event because it can cause soluble ␣-syn to form covalently linked dimers and higher M r oligomers and hence allow for fibril formation and stabilization. Our data suggest that this process is independent of Tyr residues, but the exact amino acid residues involved in this process remain to be elucidated. Thus, although the molecular mechanisms by which ␣-syn protein polymerizes and forms inclusions within cells remain elusive, the data presented here provide credence for the role of oxidative and nitrative stress in the formation of stable ␣-syn inclusions in human disease.