Reversible Inhibition of α-Synuclein Fibrillization by Dopaminochrome-mediated Conformational Alterations*

Previous studies demonstrated that α-synuclein (α-syn) fibrillization is inhibited by dopamine, and studies to understand the molecular basis of this process were conducted (Conway, K. A., Rochet, J. C., Bieganski, R. M., and Lansbury, P. T., Jr. (2001) Science 294, 1346–1349). Dopamine inhibition of α-syn fibrillization generated exclusively spherical oligomers that depended on dopamine autoxidation but not α-syn oxidation, because mutagenesis of Met, His, and Tyr residues in α-syn did not abrogate this inhibition. However, truncation of α-syn at residue 125 restored the ability of α-syn to fibrillize in the presence of dopamine. Mutagenesis and competition studies with specific synthetic peptides identified α-syn residues 125–129 (i.e. YEMPS) as an important region in the dopamine-induced inhibition of α-syn fibrillization. Significantly, the dopamine oxidation product dopaminochrome was identified as a specific inhibitor of α-syn fibrillization. Dopaminochrome promotes the formation of spherical oligomers by inducing conformational changes, as these oligomers regained the ability to fibrillize by simple denaturation/renaturation. Taken together, these data indicate that dopamine inhibits α-syn fibrillization by inducing structural changes in α-syn that can occur through the interaction of dopaminochrome with the 125YEMPS129 motif of α-syn. These results suggest that the dopamine autoxidation can prevent α-syn fibrillization in dopaminergic neurons through a novel mechanism. Thus, decreased dopamine levels in substantia nigra neurons might promote α-syn aggregation in Parkinson's disease.

Parkinson disease (PD) 1 is the most common neurodegenerative movement disorder, as it affects over one million people in North America and four million worldwide (1). PD is clinically diagnosed by four characteristic features, bradykinesia, postural instability, motor rigidity, and resting tremor. Pathologically, there is a progressive loss of dopaminergic neurons in the substantia nigra pars compacta, which results in a significant decrease in dopamine levels in the striatum followed by motor impairments in PD patients (1)(2)(3). In addition to neuron loss, intracellular proteinaceous lesions are found in different PD brain regions that are termed Lewy bodies (LBs) and Lewy neurites. LBs are found in the remaining dopaminergic neurons of the substantia nigra (4,5), but they also occur in other brainstem neurons as well as in those of the thalamus, hypothalamus, cortex, olfactory bulb, and other brain regions (4,6,7). These inclusions are now known to be comprised of filamentous polymers of ␣-synuclein (␣-syn) protein (5, 8 -14).
␣-Syn is a 140-amino acid heat-stable protein that is predominantly found in presynaptic terminals of cells of the central nervous system (15)(16)(17)(18). Studies have shown that pathological inclusions comprised of ␣-syn are found in neurodegenerative disorders other than PD, including the LB variant of Alzheimer's disease, dementia with LBs, multiple system atrophy, and related diseases collectively known as ␣-synucleinopathies (10, 12-14, 19, 20). Although ␣-syn is natively unfolded and soluble in aqueous solutions (21)(22)(23)(24)(25), conformational changes of this protein from random coil to ␤-pleated structure by unknown mechanism(s) lead to the formation of insoluble ␣-syn fibrils that are the building blocks of pathological inclusions (24, 26 -29).
Previous studies suggest that ␣-syn may play a role in the regulation of synaptic dopamine levels and its secretion. For example, down-regulation of ␣-syn in neuronal cultures results in an alteration in the number of docked vesicles (30). Faulty synaptic transmission and altered striatal dopamine levels were observed in ␣-syn knock-out mice (31,32). Furthermore, other studies have shown an interaction between the dopamine transporter and ␣-syn. Although some of the data are conflicting, these studies collectively suggest that dopamine transporter and ␣-syn form a complex that then modifies dopamine reuptake (33,34). Dopamine/␣-syn interactions also have been detected in cell culture models. For example, Xu et al. showed increased cell death in dopaminergic neurons transfected with wild type (WT) or pathologically mutant forms of ␣-syn (35), and Paxinou et al. reported the formation of ␣-syn aggregates in ␣-syn-overexpressing cells following simultaneous exposure to dopamine and nitric oxide (36).
Recently, Conway et al. showed that dopamine or L-dopa inhibits the fibrillization of recombinant ␣-syn filaments, presumably through adduct formation and stabilization of ␣-syn into "protofibrillar" structures that are incompetent to form fibrils (37). To better understand interactions between dopamine and ␣-syn, we conducted in vitro studies to determine the nature of this interaction, and we also characterized the structural and conformational properties of dopaminemodified ␣-syn. Our results demonstrate a novel mechanism whereby dopaminochrome, an oxidized product of dopamine, inhibits ␣-syn fibrillization by interacting with a specific amino acid motif in the C terminus. Remarkably, the dopamine oxidation-induced formation of oligomeric spheres that are incapable of forming mature ␣-syn fibrils is due to conformational changes rather than covalent modifications, because these changes are reversible, Thus, our novel observations could be exploited for developing better therapies for PD-related ␣-synucleinopathies.

Expression of WT, Mutant, and Truncated Recombinant Human
␣-Syn Proteins-Human WT ␣-syn cDNA was cloned into the bacterial expression vector pRK172 at the NdeI and Hind III restriction sites. Similar plasmids expressing point mutations of ␣-syn or C-terminal truncations of ␣-syn due to nonsense mutations were constructed using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) as described previously (22,38,39). WT, mutant, and truncated ␣-syn proteins were expressed in Escherichia coli BL21 (DE3) RIL cells and purified as described previously (22,40).
Analysis of Assembled ␣-Syn Proteins-␣-Syn proteins (50 l samples) were incubated at 37°C with continuous shaking at 1,000 rpm at a concentration of 5 mg/ml (345 M) in phosphate-buffered saline (PBS; 137.5 mM NaCl, 2.5 mM KCl, 10 mM Na 2 HPO 4 , and 1.75 mM KH 2 PO 4 , pH 7.5) containing 0.04% sodium azide. Dopamine was used at equimolar concentrations to ␣-syn (345 M) or one-fifth of the concentration of ␣-syn (69 M), and N-acetyl cysteine was used at 3-fold molar excess. Each sample (50 l) was overlaid with ϳ40 l of mineral oil to prevent evaporation. For sedimentation analysis, pelletable polymers of ␣-syn were collected by centrifugation at 100,000 ϫ g for 20 min. SDS sample buffer was added to supernatants and pellets, which 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 formation of amyloid polymers of ␣-syn was assayed using the K114 amyloid fluorescent dye as described previously (38,41), and ␣-syn fibrillization and/or oligomerization were monitored by transmission electron microscopy (EM) and atomic force microscopy (AFM) (see below).
Purification of Dopamine-treated ␣-Syn Proteins-WT or mutant ␣-syn proteins were exposed to equimolar concentrations of dopamine in PBS, and samples were incubated overnight at 37°C. In some experiments, byproducts of dopamine oxidation were removed from modified ␣-syn by resolving the reactants using Affi-Gel blue bead chromatography. ␣-Syn was exchanged with Microcon YM-10 centrifugal filter devices (Millipore, Bedford, MA) into 10 mM Tris, pH 7.5, and further purified through DEAE-Sepharose beads and eluted by 300 mM NaCl in 10 mM Tris, pH 7.5. ␣-Syn was exchanged into either 100 mM NaOAc, pH 7.5, or PBS. This purified ␣-syn was used for fibril assembly studies, negative staining EM, immuno-EM, AFM, circular dichroism, and Fourier transform infrared (FTIR) spectrometry.
Visualization of ␣-Syn Fibril and Spherical Oligomer Formation by EM and AFM-WT, mutant, and dopamine-modified ␣-syn samples were collected before and after incubation for EM, immuno-EM, and AFM analyses. Negative staining EM was used to observe fibrils of WT and mutant ␣-syn proteins after fibrillization. Samples were applied to 300 mesh carbon-coated grids and negatively stained with 1% uranyl acetate. A JEOL 1010 EM was used to examine these samples at magnifications up to 100,000ϫ, and images were captured with a Hermits camera (Birdgewater, MA) using software from AMT (Danvers, MA).
Purified dopamine-modified ␣-syn samples were analyzed by immuno-EM. Samples were applied onto 300 mesh carbon-coated grids, blocked with 1% bovine serum albumin in PBS, and immunostained with various primary antibodies including monoclonal antibodies that detected the N terminus (Syn505 and Syn506) and C terminus (Syn211 and Syn214) of ␣-syn and a polyclonal rabbit antibody raised to recombinant ␣-syn. Oligomers were then decorated with anti-mouse or antirabbit antibodies conjugated to 5-nm gold particles and negatively stained with 1% uranyl acetate. Control grids were stained with secondary antibody alone or with an anti-tau antibody (T14). EM grids were analyzed at magnifications up to 250,000ϫ.
A multimode atomic force microscope equipped with a J scanner (Digital Instruments, Santa Barbara, CA) was used to analyze control and dopamine-modified WT and mutant ␣-syn samples for the presence of fibrils and/or spherical oligomers. Diluted samples were applied to freshly peeled, 1 cm in diameter mica discs, and the specimens were air-dried overnight. Silicon nitride cantilevers with a spring constant of 0.06 newton/m were used for imaging, and the images were collected after the force applied to each specimen was minimized.
Circular Dichroism Spectrometry-CD spectra were recorded using a Jasc0 J-810 spectropolarimeter. Spectra were collected at 25°C in a 0.1-cm-long quartz cuvette containing the protein diluted to 0.1 mg/ml in 50 mM potassium phosphate buffer, pH 7.6.
Protein Conformational Studies by FTIR Spectrometry-An FTS 60A FTIR spectrometer (Bio-Rad) was used to analyze ␣-syn samples after the buffer was exchanged into 2 mM HEPES in D 2 O, pH 7.4, and nitrogen-dried onto a germanium internal reflection crystal. Before sample spectra were collected, rapid background scans were performed. Techniques such as smoothing, water vapor subtraction, baseline correction, and deconvolution were not performed on the raw spectra.
Synthesis of Dopaminochrome-Dopaminochrome was synthesized by dissolving dopamine at 5 mM in 10 mM sodium acetate, pH 5.8. Sodium periodate (NaIO 4 ) was added to a final concentration of 10 mM (from a stock of 100 mM in H 2 O) and incubated with constant shaking for 5 min. Reverse phase high performance liquid chromatography analysis was used to confirm that the reaction was complete and resulted in dopaminochrome as the major product.
Synthesis of Aged/Polymerized Dopamine-Dopamine (35 mM) was incubated in PBS for 96 h. Oxidized dopamine polymers were recovered by centrifugation at 200,000 ϫ g for 1 h and resuspended in PBS.
Denaturation and Renaturation of Dopamine-Modified ␣-Syn-␣-Syn was incubated with dopamine at an equimolar concentration overnight at 37°C without shaking. Dopamine-modified ␣-syn was denatured by the addition of urea to 7 M. ␣-Syn was subsequently re-natured by dialysis into decreasing concentrations of urea over a period of 48 h and eventually into PBS. Re-natured ␣-syn was then incubated with shaking at 37°C for 96 h, and the protein was analyzed by centrifugal sedimentation and K114 fluorometry.

Inhibition of ␣-Syn Protein Fibrillization by Dopamine-Re-
combinant WT human ␣-syn was incubated under conditions that result in fibril formation in the absence or presence of dopamine with or without N-acetyl cysteine (an oxidant scavenger) or tyrosine. Sedimentation analysis (Fig. 1A) and K114 fluorescence analysis (Fig. 1B), which specifically binds amyloid fibrils (41), were employed to assess the degree of ␣-syn fibrillization in these samples. Both assays revealed that WT ␣-syn does not fibrillize in the presence of dopamine, and this inhibition is prevented by the addition of the reducing agent N-acetyl cysteine to the incubation reaction (Fig. 1, A and B). Tyrosine also was used as a control because its structure is similar to that of dopamine, but it does not self-oxidize or promote oxidation and did not inhibit the fibrillization of ␣-syn fibrils (Fig. 1, A and B). Most experiments in these studies were performed with an oil overlay to prevent evaporation (see "Experimental Procedures"); however, similar results were obtained without an oil overlay (see supplemental Fig. 1, available in the on-line version of this article). Furthermore, WT ␣-syn, which was incubated with dopamine without shaking and then purified to eliminate all dopamine oxidation byprod-ucts, also did not form pelletable polymers or amyloid fibrils upon further incubation for up to 12 days in vitro (data not shown). These results indicate that dopamine oxidation is involved in the inhibition of ␣-syn filament formation.
To identify ␣-syn structures generated by dopamine inhibition, EM and AFM analyses were conducted. Although untreated WT ␣-syn produced abundant fibrils after incubation for 48 h as detected by AFM (Fig. 1C) and EM (data not shown) (23 and 40), no fibrils were observed in dopamine-treated ␣-syn samples. Instead, there was an abundant accumulation of spherical oligomers (Fig. 1D). These spheres accumulated even within 12 h of adding dopamine and no further incubation (data not shown), indicating that dopamine promoted the conversion of ␣-syn from monomers to spherical oligomers. Calculations based on AFM data suggested that these spheres are ϳ25 nm in diameter and contain ϳ25 ␣-syn molecules. When immuno-EM was conducted using monoclonal antibodies specific for the C terminus (Syn 211 and Syn 214) (42) and N terminus (Syn 505 and Syn 506) (43) of ␣-syn, immunogold labeling was detected using all of the monoclonal antibodies (Fig. 1, E and F), suggesting that both the N terminus and the C terminus of ␣-syn within these spheres are exposed. As negative controls, anti-tau or secondary antibodies alone were used to immunostain the dopamine-modified ␣-syn oligomers, and no labeling was detected (data nor shown).
Dopamine-modified ␣-Syn Spherical Oligomers Demonstrates a Predominantly Random Coil Structure with Some ␤-Pleated Sheets-CD and FTIR spectrometry were used to determine the conformation of ␣-syn after modification by dopamine. Consistent with previous observations (39,44), soluble, unassembled ␣-syn is predominantly in random coil conformation (Fig. 2, A and B) as indicated by a minimum between 197 and 200 nm in the CD spectra. Polymerized, filamentous ␣-syn is predominantly in ␤-pleated sheet conformation (Fig. 2, A and C) as indicated by the minimum of ϳ220 nm and the positive signal Ͻ195 nm in the CD spectra. Dopamine-or dopaminochrome-treated ␣-syn was demonstrated by CD analysis as being mostly randomly coiled with little ␣-helix or ␤-pleated sheet structure ( Fig. 2A), but FTIR spectrometry analysis showed that dopamine-treated ␣-syn has some ␤-pleated sheet structure.
Dopamine Inhibition of ␣-Syn Fibrillization Does Not Require Modification of Tyr, Met, and His-␣-Syn does not contain any Cys or Trp residues, known targets for dopamine modification, but the protein does contain four Met residues, four Tyr residues, and one His residue that could be affected by dopamine or products of dopamine oxidation. Thus, to investigate whether dopamine-mediated inhibition of ␣-syn fibrillization requires oxidation or other chemical modifications, ␣-syn proteins with the following mutations were generated: (i) single, double, triple, and quadruple Tyr 3 Phe mutations; (ii) single, double, and triple Met 3 Ala mutations; and (iii) His 3 Arg (H50R) mutation. The quadruple Met 3 Ala mutation could not be generated, because Met-1 is necessary for bacterial protein expression as the start codon.
Significantly, similar to WT ␣-syn, all ␣-syn mutants formed pelletable amyloid fibrils, and this process was inhibited by dopamine (Fig. 3, A and B). To confirm that dopamine arrested fibrillization of mutant ␣-syn at the spherical oligomeric stage, dopamine-treated Tyr 3 Phe ␣-syn with four types of mutations, Met 3 Ala ␣-syn with three types of mutations, and H50R ␣-syn were examined by AFM. Spherical oligomers with similar morphology and size to dopamine-treated WT ␣-syn accumulated with all mutant proteins (data not shown). Taken together, our data demonstrate that dopamine inhibition of ␣-syn fibrillization is independent of single or multiple Tyr, Met, and His residues.
To demonstrate directly that ␣-syn YEMPS residues mediate the inhibitory effect of dopamine, an ␣-syn mutant (designated 125-129) was generated wherein all five 125-129 amino acids residues were substituted to FAAFA, respectively. The effects of dopamine (at both equimolar and a sub-stoichiometric concentrations, i.e. 1:5 ratio of dopamine to protein) on WT ␣-syn, A53T ␣-syn, and ␣-syn 125-129 fibrillization were compared. At both concentrations, dopamine inhibited fibrillization of WT ␣-syn (Fig. 4, B and C). Equimolar concentrations of dopamine also inhibited fibril formation of A53T ␣-syn, but at a ratio of 1:5 A53T ␣-syn was able to partially fibrillize. In contrast, the ␣-syn 125-129 mutant was more resilient to dopamine-mediated inhibition of fibrillization at either concentration, although some inhibition was still observed. These data were confirmed with EM analysis that showed abundant mutant ␣-syn 125-129 fibrils in the presence of dopamine (Fig. 4D). Taken together, these results establish that residues 125-129 in ␣-syn are involved in dopamine inhibition of ␣-syn fibrillization.
Dopaminochrome, a Product of Dopamine Oxidation, Is Responsible for the Inhibition of ␣-Syn Fibril Formation-To determine whether byproducts of dopamine autoxidation may inhibit ␣-syn filament formation rather than dopamine itself, the effects of dopaminochrome, a relatively stable oxidized product of dopamine, was analyzed for WT and mutant ␣-syn 125-129. Significantly, freshly synthesized dopaminochrome inhibited ␣-syn fibrillization almost as effectively as dopamine (Fig. 5, A and B). However, the mutant ␣-syn 125-129 protein can fibrillize in the presence of dopaminochrome. Furthermore, dopamine that has been allowed to oxidize and polymerize over time had little effect on the fibrillization of ␣-syn (Fig. 5C), indicating that oxidized intermediates such as dopaminochrome are responsible for the inhibition of ␣-syn filament formation.
Reversibility of Dopamine Inhibition of Filament Formation-The ability of dopaminochrome to inhibit filament formation suggests that this process may not be due to a covalent modification of ␣-syn. Consistent with this notion, mass spectrometry analyses of isolated dopamine-treated ␣-syn did not reveal significant dopamine/␣-syn adducts, and [ 3 H]dopamine incorporation analyses showed that Ͻ0.1% of the ␣-syn molecules were adducts of dopamine (data not shown). The lack of dopamine/␣-syn adducts suggests that oxidized dopamine may induce conformational changes in ␣-syn, resulting in proteins that are fibrillization-incompetent. Moreover, these findings imply that this inhibition should be reversible. To test this hypothesis, dopamine-treated ␣-syn spherical oligomers were denatured in 7 M urea and re-natured by dialysis in PBS. Remarkably, this treatment led to the near complete recovery of the ability of dopamine-treated ␣-syn to assemble into fibrils (Fig. 5, D and E), which were ultrastructurally identical to those formed by untreated ␣-syn (Fig. 5F).

DISCUSSION
The studies presented here provide novel mechanistic insights into how dopamine inhibits ␣-syn fibrillization predominantly by inducing alterations in protein conformation that affect ␣-syn polymerization. Consistent with previous reports (37), our data show that dopamine prevents ␣-syn from forming mature amyloid fibrils and that oxidation is needed for this inhibition. However, our findings indicate that dopamine-induced covalent modification of ␣-syn is not required to prevent filament formation. Rather, inhibition of filament formation is due to the formation of oxidized dopamine by-products that induce conformational changes in ␣-syn to form species that are unable to proceed to assemble into mature fibrils.
Several key pieces of evidence support these conclusions. First, using mass spectrometry we and others (46) were unable to detect significant amounts of ␣-syn dopamine adducts, which is consistent with findings that dopamine can inhibit ␣-syn filament formation at sub-stoichiometric concentrations (37). Second, under the conditions used here that inhibit ␣-syn fila- ment formation, only a minute amount of ␣-syn was covalently modified by dopamine (Ͻ0.1%), as demonstrated by using [ 3 H] dopamine in radiolabeling studies (data not shown). Third, mutagenesis studies of the major amino acid residues that could be substrates for covalent modification by dopamine failed to show any effect. Although treatment with dopamine did lead to methionine oxidation as observed by the presence of species with an increased molecular mass of 64 Da using mass spectrometry (data not shown), this oxidation was not responsible for inhibition of ␣-syn fibrillization as demonstrated by studies of Met-to-Ala ␣-syn mutants and by previous studies (38). Finally, ␣-syn filament formation could be inhibited by the addition of a by-product of dopamine oxidization, i.e. dopaminochrome, although other intermediates may also be involved. It is noteworthy that fully oxidized/polymerized dopamine is ineffective in inhibiting ␣-syn fibril formation, and this failure of polymerized, oxidized dopamine to interact with ␣-syn might be due to chemical inactivation or steric hindrance.
The data presented here also support the notion that dopamine by-products, such as dopaminochrome, interacting with the amino acid motif YEMPS in ␣-syn are involved in the inhibition of filament formation. These dopamine intermediates appear to act as molecular chaperones, inducing conformational changes in ␣-syn as observed by FTIR spectrometry and the accumulation of structural "spheres" observed by AFM. It has been proposed that ␣-syn spherical oligomers generated in the absence of dopamine or dopaminochrome may be intermediates in the pathway that lead to mature fibril formation (44,49). However, dopamine-induced ␣-syn spheres could be structural variants that are on an "off-filament pathway" and therefore unable to progress to form fibrils. Significantly, this altered form of ␣-syn can be rendered fibrillization-competent by denaturation/renaturation, consistent with the role of conformational alterations rather than covalent modifications in dopamine inhibition of ␣-syn fibrillogenesis.
These finding may have important implications for the role of dopamine oxidation in the pathogenesis of PD and future PD drug discovery efforts. It is hypothesized that dopaminergic neurons may be selectively vulnerable in PD because of their increased exposure to oxidation stress as a result of dopamine biochemistry. However, our data here suggest that intermediates in dopamine oxidation may prevent the formation of ␣-syn filaments, although other catecholamines might have similar effects (37,46). Nevertheless, it remains to be determined if spherical oligomers generated because of dopamine oxidation have any toxic effects. It also is difficult to assess how many oxidized dopamine intermediates exist in cells at steady state, because it is likely that these compounds readily convert to neuromelanin. It is also interesting to speculate that a reduction in dopamine levels may occur early in PD and result in an increased propensity for WT ␣-syn to form fibrillar inclusions, but this may partially be countered by L-dopa, which has a similar effect on ␣-syn filament formation as dopamine (37,46). Although LB-like inclusions with mature ␣-syn fibrils developed in multiple brain regions in animal models of synucleinopathies (47,48), dopaminergic neurons such as those in substantia nigra are remarkably spared and devoid of LB-like inclusions, suggesting that normal dopamine levels may prevent ␣-syn fibrillogenesis in vivo.
In summary, this study provides evidence for an alternative polymerization pathway for ␣-syn that does not culminate in mature ␣-syn fibrils. Furthermore, we show that ␣-syn fibrillization can be affected by physiological, molecular chaperones that may have important consequences for disease pathogenesis. These findings provide a framework for further studies to define the physiological role of dopamine on ␣-syn inclusion formation, and our observations could be exploited to develop novel therapies for PD and related ␣-synucleinopathies.