• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Nov 2, 2010; 107(44): 18850–18855.
Published online Oct 14, 2010. doi:  10.1073/pnas.1012336107
PMCID: PMC2973859
Biophysics and Computational Biology

Identification of a helical intermediate in trifluoroethanol-induced alpha-synuclein aggregation

Abstract

Because oligomers and aggregates of the protein α-synuclein (αS) are implicated in the initiation and progression of Parkinson’s disease, investigation of various αS aggregation pathways and intermediates aims to clarify the etiology of this common neurodegenerative disorder. Here, we report the formation of short, flexible, β-sheet-rich fibrillar species by incubation of αS in the presence of intermediate (10–20% v/v) concentrations of 2,2,2-trifluoroethanol (TFE). We find that efficient production of these TFE fibrils is strongly correlated with the TFE-induced formation of a monomeric, partly helical intermediate conformation of αS, which exists in equilibrium with the natively disordered state at low [TFE] and with a highly α-helical conformation at high [TFE]. This partially helical intermediate is on-pathway to the TFE-induced formation of both the highly helical monomeric conformation and the fibrillar species. TFE-induced conformational changes in the monomer protein are similar for wild-type αS and the C-terminal truncation mutant αS1-102, indicating that TFE-induced structural transitions involve the N terminus of the protein. Moreover, the secondary structural transitions of three Parkinson’s disease-associated mutants, A30P, A53T, and E46K, are nearly identical to wild-type αS, but oligomerization rates differ substantially among the mutants. Our results add to a growing body of evidence indicating the involvement of helical intermediates in protein aggregation processes. Given that αS is known to populate both highly and partially helical states upon association with membranes, these TFE-induced conformations imply relevant pathways for membrane-induced αS aggregation both in vitro and in vivo.

Keywords: amyloid, circular dichroism, Parkinson’s disease, principal component analysis, protein misfolding

Parkinson’s disease (PD) is one of a number of synucleopathies in which aggregation of the protein α-Synuclein (αS) is linked to pathogenesis (1). αS is intrinsically disordered, but in the presence of lipid or detergent vesicles or micelles, adopts a highly helical structure in which its N-terminal region is membrane-bound and the C-terminal tail remains predominantly free and unstructured (2, 3). Although most PD cases are sporadic or idiopathic, three point mutations of α-Synuclein– A53T, A30P, and E46K– are associated with familial and early-onset disease (see Review (4)).

In addition to its free and membrane-bound states, αS adopts partially structured intermediate conformations under low-pH or high-temperature conditions (5). A folding intermediate has also been detected at low [TFE] (6). Conditions favoring the formation of these intermediates also promote amyloid fibril growth, possibly implicating intermediate conformers as key species in the aggregation pathways.

Here, we examine TFE-induced monomer conformational changes, oligomerization, and fibrillization in detail for wild-type (WT) αS, C terminally truncated WT αS (αS102), and the PD-associated αS mutants A30P, A53T, and E46K, expanding upon previous studies by Munishkina, et. al. (6) and Li, et. al. (7). This research also complements our previous fluorescence correlation spectroscopy studies of αS membrane interactions (8) and protein equilibrium structural dynamics (9).

Helical intermediates have been reported to promote fibril formation of a number of amyloidogenic proteins (10). We show that αS is likely to aggregate via such an intermediate in the presence of TFE, suggesting that membrane-induced αS aggregation may also involve the formation of a helical intermediate. Furthermore, TFE-induced fibrils are β-sheet rich and resemble previously reported aggregates formed by C terminally truncated αS (11), as well as structures induced by detergent and lipid interactions (12, 13), which may be linked to PD initiation and progression (1417).

Results

Ultrastructure of TFE-Induced αS Aggregates.

Fig. 1 AD shows micrographs of various aggregates formed from WT αS after two weeks in a shaking incubator at 37 °C. Typical long, rigid amyloid fibrils form at 0% TFE (all percentages v/v). At 5% TFE, a combination of typical amyloid fibrils and shorter, flexible fibrillar structures are observed. When [TFE] is increased to 10%–15%, classic amyloid fibrils disappear and only the flexible, short “TFE fibrils” are observed.

Fig. 1.
WT αS aggregate characteristics as a function of [TFE]. (AD): TEM micrographs of structures grown from 50 μM WT αS, after 2 wk incubation at 37 °C with shaking in the presence of (A) 0%, ...

Aggregate production for the samples in Fig. 1 AD was quantified by centrifugation, UV absorbance, and thioflavin-T fluorescence enhancement (Fig. 1E). The amount of aggregate produced rises sharply at ≥10% TFE where TFE fibrils predominate.

Additional images of TFE-induced WT αS aggregates grown in a variety of solution conditions show that TFE fibrils can be grown at 25 °C even in the absence of shaking when [TFE]≥ ~ 10% (Fig. S1A). We have not observed classic amyloid fibril formation in the absence of shaking after incubations of up to 3 wks in 0%–20% TFE. Ultrastructurally, TFE fibrils appear to be helical, with a strand width of ~11 nm (Fig. S2A). The overall fibril diameters are ~11–20 nm and appear to vary due to stretching or compression of the helical winding, while the minimum thickness of the strands is ~5–6 nm. Amyloid fibrils in our 0% TFE sample range in diameter from ~9–23 nm (the mean width is ~12 nm) and are thus similar in width to TFE fibrils, but are much longer and straighter. Structures that resemble closed, distorted rings can sometimes be found via TEM (Fig. 1C and Fig. S2 B and C). Rings were most common in samples that were incubated at 37 °C with shaking, although we also observed them in some room-temperature samples (e.g., Fig. S2C). It is not clear whether ring-like structures are actually closed loops or whether their appearance is accidental due to fibril flexibility and artifacts of drying onto the TEM grids. The extent of aggregation and thioflavin-T fluorescence emission enhancement varies as a function of [TFE] for αS samples incubated at room temperature under quiescent conditions (Fig. S1B), with TFE fibril production occurring above ~10% TFE, and associated with some thioflavin-T binding. Similar TFE fibril production behavior at 25 °C is observed for the A30P, A53T, and E46K PD-associated αS mutants (Fig. S1B). Qualitatively, the amount of aggregate produced is highest for the E46K mutant, while the extent of A30P aggregation is reduced. In addition, the C-terminal truncation mutant αS102 forms fibrillar aggregates when incubated in ~10%–15% TFE (Fig. S2 EF), indicating that the N-terminal portion of the protein is sufficient for TFE-induced fibril formation.

TFE-Induced Secondary Structural Changes in Monomeric αS.

The far-UV CD spectrum of dilute (0.5 μM) WT αS in 0% TFE (Fig. 2A) is relatively flat except for a deep minimum around 198 nm, consistent with a highly disordered protein. As [TFE] increases, the signal at 218–222 nm decreases, reflecting increased amounts of secondary structure. At 60% TFE, the spectrum has minima at 208 and 222 nm, indicating that αS adopts a highly α-helical conformation under these conditions. Surprisingly, sets of spectra for 0%–14% and 17%–60% TFE share isodichroic points (Fig. 2A insets), consistent with coexistence of two distinct secondary structural conformations in each range of TFE concentrations. We have verified that the curves in Fig. 2A are representative of monomeric protein by examining time- and concentration-dependent variations in the spectra (see SI Text and Fig. S3).

Fig. 2.
Secondary structural changes induced by TFE for WT αS at 25 °C. (A) Far-UV CD spectra for 0.5 μM WT αS variant in 0%–60% TFE. The TFE concentrations for spectra with increasing negative ellipticity ...

By plotting the mean residue ellipticities at 198 nm vs. 222 nm ([θ]222 vs. [θ]198), we can construct a transition diagram (18) from the CD spectra in Fig. 2A, enabling identification of structural transitions for the monomer protein (Fig. 2B). Points derived from spectra corresponding to shifts in an equilibrium between two conformations appear as straight lines in this diagram; the spectra that share isodichroic points in Fig. 2A form straight lines in our transition diagram. Our observation of two different, adjacent linear segments indicates that αS is sampling at least three secondary structure conformations: an unfolded conformation (U), which is approximated by the 0% TFE point; a well-folded, α-helical species (F), which is most similar to the 60% TFE point; and an intermediate secondary structural conformation (I) that is populated at moderate [TFE]. Along the low-TFE line (bottom right of the plot), the U and I conformations coexist, while the I and F states are populated along the high-TFE line (top left).

Similar structural transitions are observed for αS102, as well as the A30P, A53T, and E46K αS mutants, with CD spectra showing a progression from a disordered conformation to α-helical secondary structure with increasing [TFE], and low- and high-[TFE] curves sharing isodichroic points (Fig. S4). The qualitative behavior of the αS102 mutant is very similar to WT, although the overall magnitude of the mean residue ellipticity for αS102 mutant is increased at moderate- to high-TFE (Fig. S4 A and E, S5 A and B), indicating that a larger fraction of residues adopt secondary structure in the truncation mutant. Moreover, when transition diagrams are constructed for the three PD-associated mutants, their coexistence lines are nearly identical to those of WT αS (Fig. S4 FH, S5E). Therefore, the TFE-induced folding landscapes of the αS variants contain comparable structural transitions. Surprisingly, plots of [θ]222 and [θ]198 as a function of [TFE] for all the αS variants (Fig. S5 AD) appear superficially to be sigmoidal, which is likely due to the fact that the intermediate spectrum [θ]222 and [θ]198 values lie in between the values for the U and F states. Thus, the transition diagrams reveal information about intermediate states that is hidden in these plots.

Table S1 shows the TFE concentrations and isodichroic points that correspond to U↔I and I↔F coexistence for all five αS variants. Note that, for a set of CD curves that share an isodichroic point, the ellipticity of all conformations that contribute to the spectra is the same at the isodichroic. Therefore, the I state ellipticity is equal to the values measured at these points. Additionally, the point at which the two straight lines in the transition diagrams intercept should correspond to the CD values for the pure I state; these values are also shown in Table S1.

Reconstructed I State Spectrum.

Table S1 shows ellipticity values for the pure I state at four points. However, a spectrum that covers a larger wavelength range is desirable to obtain information about the secondary structure of this conformer; therefore we reconstruct the I state curve for 195–260 nm using two methods. We use Principal Component Analysis (PCA) to reduce the dimensionality of our datasets, and then find the I state via fits (in the new coordinate system) to points whose spectra share isodichroics (SI Text and Fig. S6). In addition, we use the information from Table S1, along with Maximum Likelihood Estimation (MLE), to find fractions of the U and F states as a function of [TFE], and then subtract these contributions from our measured spectra to reconstruct the I state spectrum (SI Text). Results for both methods for WT αS are shown in Fig. 3A. The spectra obtained from both methods are similar to one another and are also consistent with the values in Table S1. The PCA results for all the αS mutants are shown in Fig. 3B; the MLE estimates are similar to the PCA results and are shown in Fig. S6 DG.

Fig. 3.
Reconstructions of the I state spectra. (A) Predicted spectra for WT αS. The solid lines show the spectrum calculated via PCA. The dashed lines show results of MLE analysis, which were calculated using spectra that shared the low-TFE isodichroic ...

The reconstructed I state spectra for all the αS variants are similar to one another in that they all exhibit a minimum around 222 nm that is suggestive of α-helical secondary structure (Fig. 3B). However, the accompanying minimum expected for pure α-helical structure at 208 nm is shifted to slightly shorter wavelengths. In addition, the magnitude of the signal at 222 nm is less than would be expected for fully helical structure for all five variants, indicating that the intermediate state is partially unstructured in all cases.

Using [θ]222, we estimate the fractional helicity (19) for the I and F states (Table 1). The percent helicity of both states is higher for αS102 compared to WT, but the number of residues in a helical conformation is similar, indicating that the N terminus of full-length αS is likely to be the region adopting structure in the presence of TFE. For all the variants, ~24 residues are predicted to be in a helical conformation for the I state and ~86 residues are helical in the F state. Notably, helicity is slightly lower for the A30P I state. Deconvolution analysis of the CD spectra indicated that TFE does not lead to a significant increase in β-sheet content between the U and I states (see SI Text).

Table 1.
Estimates of the percentage of α-helical structure (19) for αS in the pure I and F states, based on the magnitude of the ellipticity at 222 nm.

Conformer Populations for Monomeric αS.

The experimental CD data at all TFE concentrations are well fit by linear combinations of the reconstructed I state spectra, and the 0% and 60% TFE spectra (Fig. 4A and Fig. S7). The corresponding populations (Fig. 4B), which are very similar for all five αS variants, show that the U state becomes depleted in favor of the I state as [TFE] is increased to ~15%, while at higher TFE, the F state population increases. The I state appears to be an intermediate in the TFE-induced conversion of U to F and in addition is significantly populated between ~10% to ~20% TFE, where TFE fibril formation is maximal.

Fig. 4.
Calculations based on linear combinations of the pure U, I, and F states. (A) Fit results (black lines) for WT αS CD spectra (open circles, data as in Fig. 2A), where the fitted curves were calculated from linear combinations of the 0%, ...

Secondary Structure of TFE-Induced αS Oligomers and Fibrils.

At protein concentrations of 2 μM in solutions that contain intermediate (~12%–20%) amounts of TFE, the CD spectra change over time as oligomerization occurs, enabling an analysis of both secondary structure changes and kinetics (Fig. 5). The initial CD curves have double minima near 205 and 220 nm, and are consistent with partially unstructured protein. As time passes, a single minimum appears near 216 nm, signifying the formation of β-sheet-rich structure. Because the appearance of β-structure is both concentration- and time-dependent, we believe it reflects the formation of oligomeric species. Interestingly, the curves share isodichroic points at ~210 nm, suggesting that the systems are undergoing all-or-nothing transitions between two states. However, because oligomers with similar secondary structure could result in nearly identical far-UV CD curves, we may not be able to resolve conversions between oligomeric states, such as the association of smaller oligomers into larger species. We do not observe fibrils via TEM for samples incubated for ≤ 6 h at 25 °C, even for αS concentrations as high as 50 μM; therefore we believe that the CD spectra changes for Fig. 5 correspond to the formation of nonfibrillar oligomeric species.

Fig. 5.
Oligomer formation kinetics 15% TFE. (A) Far-UV CD spectra taken at various time points for 2 μM WT, A30P, A53T, and E46K αS in 15% TFE at 25 °C. The initial time point for each plot has the least negative [θ] ...

Measurements at a higher protein concentration of 5 μM show that the appearance of the ellipticity minimum at 216 nm is nearly complete within the 1–3 min mixing time for WT and the PD-associated mutants (excepting A30P), further confirming the concentration dependence of the initial oligomerization reaction (Fig. 5B). Fits of the data to a single exponential model result in apparent rate constants kapp (see SI Text). Visual inspection of the data in Fig. 5 A and B, in combination with the fit results (Table S2), reveals that the E46K mutant reaches steady-state fastest, while A30P is slowest. Therefore, oligomerization rates appear to follow a similar series (A30P < WT ≤  ≈ A53T < E46K) as the extent of aggregation data (Fig. S1B), which indicated that the A30P mutant forms fibrils least readily and aggregation production is highest for E46K.

CD spectra for 50 μM WT αS solutions that were incubated for two weeks at room temperature (Fig. S8A) demonstrate that mature TFE fibrils are also rich in β-strand structure. The presence of such fibrils in these solutions was confirmed by TEM (Fig. S8 B and C). Although the fraction of TFE fibrils was not measured for these samples, the data in Fig. S1B, obtained for solutions incubated under identical conditions, indicates that a significant fraction of protein is incorporated into large aggregates. The ellipticity at 216 nm for samples containing TFE fibrils is within experimental error of the values measured for the rapidly formed oligomers, indicating that early- and late- aggregates contain similar secondary structure. The spectra are also similar to those obtained for typical amyloid fibrils (2022).

Discussion

We have investigated TFE-induced conformational changes, oligomerization, and fibril production for WT human αS, C terminally truncated αS, and three PD-associated αS variants. Our results demonstrate that the TFE-induced folding landscapes for the mutants are nearly identical to WT, but the kinetics of the oligomerization process vary among the disease-associated mutants. An intermediate conformational state, which has a far-UV CD spectrum that is consistent with the presence of significant α-helical structure, is highly populated at TFE concentrations where TFE fibril production is maximized. By examining the αS102 mutant, we verify that TFE-induced conformational changes involve the N-terminal portion of the protein.

TFE Induces Short, Flexible Fibrils.

CD data (Fig. S8A) indicate that TFE-induced fibrils are β-sheet-rich, suggesting that they may represent a type of amyloid aggregate. TFE fibrils also exhibit a degree of thioflavin-T binding (Fig. 1E). Nonetheless, our current data do not suffice to unequivocally establish whether these fibrils are a form of amyloid; only X-ray or electron diffraction experiments will be able to determine whether these species contain “cross-β” structure.

When imaged via TEM, TFE-induced fibrils have a flexible helical ultrastructure (Fig. 1, Figs. S1S2). As far as we know, these structures have not been extensively studied by TEM, although we found some images of possibly similar aggregates in the literature (1113). Crowther, et. al. (11) show micrographs of both typical amyloid fibrils and irregular fibrillar structures, which appear similar to TFE-induced species, for 1-120 C terminally truncated αS and also report “small irregular wavey assemblies” formed from 1–130 truncation. In addition, Broersen, et. al. (12) report images of aggregates induced by incubation of αS in the presence of the polyunsaturated acids arachidonic acid and docosahexaenoic acid which are qualitatively similar to TFE fibrils. Also, while this paper was in review, a study was published which described detergent-induced formation of species that may be similar to our TFE fibrils (13). Although more research must be done to evaluate whether species produced by truncation mutations and/or lipid and detergent interactions are indeed related to TFE fibrils, the potential similarities with previously observed structures are particularly important in light of the hypothesis that intermediate or alternative oligomeric or fibrillar species are responsible for PD toxicity (23), recent findings that C-terminal truncation of αS can lead to neuron loss and increased susceptibility to stress in transgenic mouse models (14, 15), and multiple lines of evidence that potentially link lipid interactions and metabolism with PD etiology (16).

The “ring-like” structures we observed via TEM (Fig. 1C and Fig. S2 B and C) may also be similar to annular structures imaged using atomic force microscopy (AFM) (23, 24). However, difficulties in comparing widths measured via TEM to heights measured via AFM prevent us from definitively verifying that these structures are comparable.

TFE Induces a Partially Helical, Monomeric Intermediate.

In contrast to previous studies by Li et al. (7) and Munishkina, et al. (6), we investigated TFE-induced structural transitions in monomeric αS by examining relatively dilute solutions (0.5 μM compared to ~14 μM in Li et al. and ~35 μM in Munishkina, et al.). The higher concentrations used in these previous studies led to the conclusion that TFE stabilized an intermediate containing primarily β-sheet structure. Here, we demonstrate instead that the monomer protein samples three distinct conformations: an unfolded state, a partially structured intermediate, and a well-folded α-helical conformer (Fig. 2, Fig. 3). Increased population of the α-helical intermediate state is correlated with formation of the β-sheet-rich, short flexible fibrils (Fig. 4B). It is possible that changes in solution conditions may favor both structure formation and aggregation via separate mechanisms. However, because TFE fibril growth is strongly correlated with increasing population of the intermediate state for all WT αS variants, the simplest explanation implied by our data is that the intermediate conformer is on-pathway to TFE fibril formation, although such an assertion is very difficult to prove (10).

The TFE-induced structural intermediate we observe can be compared to previously reported αS folding intermediates. Qualitatively, our low-TFE CD spectra are similar to data reported for WT αS at high temperature and low pH (5). However, recent NMR studies show that decreased pH in fact leads to an increase in helical structure in the C-terminal tail of αS (25), while our data for the αS102 truncation mutant show that TFE-induced structural changes involve the N-terminal portion of αS. Thus, the TFE and low-pH intermediates differ in regards to the location of secondary structure. A more detailed comparison with the high-temperature state awaits further characterizations.

Interestingly, far-UV CD spectra of detergent micelle-bound and membrane-bound αS show a high degree of α-helical structure and appear similar to our F state curves (26, 27). The number of residues predicted to be in a helical conformation based on our CD data here (~86) is in good agreement with the number of residues that are known to be helical in the micelle- and membrane-bound structures (2629). In addition, our results for the C-terminal truncation mutant show that TFE-induced structural changes involve the N-terminal portion of αS, which is consistent with data for membrane- and micelle- bound conformations. If the TFE-induced F state does correspond to the highly helical membrane-bound state, then the partially helical I state, which we show is on the folding pathway to the F state, may also have a corresponding membrane-associated intermediate that could potentially be involved in membrane-induced aggregation in vivo. Indeed, evidence exists for such an intermediate; both ESR studies and recent NMR studies have reported observations of partially helical membrane-bound states of αS (30, 31). Furthermore, the N terminus of αS contains a region with an elevated intrinsic helical propensity, which was proposed to nucleate helix formation upon membrane-binding by the protein (3, 27). The length of this region was estimated to be around 32 residues (3), which is fairly similar to our estimate of the number of helical residues (~24) in the TFE-induced I state, suggesting that the I state may be comprised of helical structure in this region. The slight drop in helical content of the intermediate for the A30P mutant (Table 1), which falls within this region, provides further support for this idea, although direct correspondence between membrane-associated conformations and the TFE-induced I state cannot be establish based on our current data.

Although it was previously known that membranes or detergents can facilitate the aggregation of αS (32, 33), and the protein is highly helical when bound to membranes or detergents as a monomer (3, 2629, 33), it has never previously been shown, to the best of our knowledge, that any helical αS conformations are directly involved in inducing the aggregation of this protein. Ahmad et. al. (22) found that submicellar detergent concentrations induced a partially helical ensemble of αS that was correlated with fibril elongation. However, a discrete helical intermediate was not identified and their conditions could not support fibril formation in the absence of seeding. Past studies of TFE effects on αS structure, conducted at higher [αS], identified β-sheet-rich intermediates, likely corresponding to the rapidly formed oligomers that we observe, which obscured details of the helical I state conformation (6, 7).

The association of α-helical intermediates with amyloid fibril formation has been documented for a number of different amyloidogenic proteins or peptides, including the Aβ peptide and IAPP (10). Our demonstration that TFE induces a significantly helical intermediate conformation of αS, which is strongly associated with fibril formation, adds αS to the list of proteins that aggregate via helical intermediates, at least under some conditions. The mechanism by which β-sheet-rich aggregates form from α-helical intermediates is currently unclear. One possibility involves helix-helix interactions leading to alignment of unstructured regions adjacent to helical segments, enabling oligomerization followed by β-sheet formation and propagation (10).

PD Mutations Alter TFE-Induced Aggregation Kinetics, but Not Monomeric Structural Transitions.

We find that the TFE-induced folding landscapes for the A30P, A53T, and E46K mutants are nearly identical to WT αS, which is in accordance with previous research that showed that the pH- and temperature- induced secondary structural conformations are similar for A30P, A53T, and WT αS (34). All three mutants have also been observed to undergo similar structural transitions to WT αS in the presence of detergents or lipids (3537), although the A30P mutation may lead to a slight local reduction in helical structure. Thus, secondary structural transitions appear to be largely similar among αS mutants. In contrast, oligomerization and fibrillization behavior vary significantly between PD-linked mutants, with amyloid fibril formation rates observed in the order A30P < WT < A53T/E46K (23, 35). Likewise, we find that TFE-induced oligomerization rates vary significantly among the αS variants despite their nearly identical monomer secondary structure landscapes. Interestingly, the rates of TFE-induced fibril formation (A30P < WT < A53T < E46K) follow the same order as that observed for amyloid formation in the absence of TFE, suggesting that similar properties and effects may be controlling aggregation kinetics in both cases.

Whatever the effects of the PD mutations may be, they do not appear to significantly alter the TFE-induced conversion of the disordered free state to the TFE-induced intermediate. Thus, their effects may become important subsequent to this step, either during the initial formation of oligomeric species from the monomeric intermediate or during subsequent interconversions among oligomers and fibrillar species. An additional potential effect of disease-linked mutations may be to favor some aggregation pathways over others, rather than to accelerate a specific step during a single pathway. The existence of multiple types of fibrillar aggregates is clearly demonstrated both here and in previous studies (11, 13, 20, 21, 38) but their relationship to each other, the degree of overlap in their formation pathways, and the influence of mutations on which type of aggregate is formed remain unclear at present and will require further investigation. In particular, it is unclear whether the TFE-induced αS intermediate is on-pathway to the formation of classical amyloid fibrils, in addition to TFE fibrils.

In conclusion, we have shown that intermediate concentrations of the fluorinated alcohol TFE led to rapid aggregation of the PD-linked protein αS into short fibrillar β-sheet aggregates. TFE-induced fibril formation is most efficient under conditions that cause residues in the N-terminal portion of monomeric αS to populate a partially helical intermediate state, which is therefore likely to be on the pathway to TFE fibril formation. To our knowledge, this evidence is the best to date for an αS aggregation pathway that involves a helical intermediate, and adds to indications that helical intermediates may be generally important in amyloid aggregation pathways. We propose that the TFE-induced αS intermediate may resemble membrane-associated conformations; therefore the TFE-induced aggregation pathway may be related to pathways of membrane-induced aggregation, which could be significant in vivo, where αS is known to bind to synaptic vesicles and possibly other membrane surfaces (16). We demonstrate that the formation of the TFE-induced intermediate is not significantly affected by any of the three PD-linked αS mutations, but that all three mutations do influence the overall rate of TFE fibril formation, indicating that the mutations exert their effects subsequent to the formation of the intermediate state. TFE-induced fibrils are ultrastructurally similar to species detected for αS 1–120 and 1–130 truncation mutants (11), and may be related to aggregates produced by interactions between αS and lipids and detergents (12, 13), potentially indicating that TFE fibrils may be relevant for understanding PD progression.

Materials and Methods

All solutions were buffered with 10 mM pH 7.5 sodium phosphate. Recombinant WT and mutant αS were produced and purified as previously described (39). Lyophilized αS variants were solubilized by dissolving at 1–2 mg/mL in buffer, followed by filtration through a 100 kDa (Microcon YM-100) centrifugal spin filter (Millipore).

Fibrils were grown by incubating 50 μM of αS variants for 14 d in solutions containing 0%–15% TFE. Sodium azide (0.02% w/v) was added to the solutions as a preservative. After incubation, samples were fractionated via centrifugation at 16,000 x g for 1 h. UV absorbance at 275 nm of ~10-fold dilution of the supernatant fraction was used to quantify the amount of soluble protein (which may include small oligomers) present in the samples after aggregation. The aggregated fraction was diluted into a buffered, 20 μM thioflavin-T solution. Fluorescence emission spectra were measured using an excitation wavelength of 460 nm. Signals were compared by integrating the emission spectra from 475–485 nm, subtracting the baseline (20 μM thioflavin-T in buffer) emission from the sample value, and normalizing to the baseline, resulting in the “enhancement factor” by which the sample peak intensity exceeds the baseline value.

Far-UV CD data were obtained using a 1 nm bandwidth. Buffer-only baseline samples were measured and subtracted from all spectra and a noise-reducing option in the instrument software was used to smooth the data. Scan speeds of 1–2 s per nanometer were used, (see SI Text).

EM images were obtained with negative-staining TEM. A 5–10 μL droplet of a sample solution was placed onto a freshly glow-discharged, carbon-coated formvar, copper grid. After 2 min, the sample solution was wicked off with filter paper, the grid rinsed with deionized water, and a 5 μL droplet of 2% (w/v) phosphotungstic acid stain (pH 7) placed on the grid. After 1 min, the staining solution was wicked away and the grid air dried.

Supplementary Material

Supporting Information:

Acknowledgments.

We thank J. Grazul, Y. Zhang, and D. Muller for invaluable help with EM imaging and analysis, J. Sethna for help with data analysis, and W. Zipfel, B. Williams, M. Williams, and H. Chen for assistance. This research made use of the Hudson Mesoscale Processing facility of the Cornell Center for Materials Research (CCMR) with support from the National Science Foundation (NSF) Materials Research Science and Engineering Centers (MRSEC) program (DMR 0520404), and was supported by National Institutes of Health (NIH) Grants AG026650 (to W.W.W.), AG019391 and AG025440 (to D.E.), the NSF Science and Technology Center program under agreement No. ECS-9876771 (to V.L.A.) the Irma T. Hirschl Foundation (to D.E.), and a gift from Herbert and Ann Siegel (to D.E.)

Footnotes

The authors declare no conflict of interest.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1012336107/-/DCSupplemental.

References

1. Spillantini MG. Parkinson’s disease, dementia with Lewy bodies and multiple system atrophy are alpha-synucleinopathies. Parkinsonism Relat D. 1999;5:157–162. [PubMed]
2. Davidson WS, Jonas A, Clayton DF, George JM. Stabilization of alpha-synuclein secondary structure upon binding to synthetic membranes. J Biol Chem. 1998;273:9443–9449. [PubMed]
3. Eliezer D, Kutluay E, Bussell R, Jr, Browne G. Conformational properties of alpha-synuclein in its free and lipid-associated states. J Mol Biol. 2001;307:1061–1073. [PubMed]
4. Lesage S, Brice A. Parkinson’s disease: from monogenic forms to genetic susceptibility factors. Hum Mol Genet. 2009;18:R48–59. [PubMed]
5. Uversky VN, Li J, Fink AL. Evidence for a partially folded intermediate in alpha-synuclein fibril formation. J Biol Chem. 2001;276:10737–10744. [PubMed]
6. Munishkina LA, Phelan C, Uversky VN, Fink AL. Conformational behavior and aggregation of alpha-synuclein in organic solvents: Modeling the effects of membranes. Biochemistry. 2003;42:2720–2730. [PubMed]
7. Li HT, Du HN, Tang L, Hu J, Hu HY. Structural transformation and aggregation of human alpha-synuclein in trifluoroethanol: non-amyloid component sequence is essential and beta-sheet formation is prerequisite to aggregation. Biopolymers. 2002;64:221–226. [PubMed]
8. Rhoades E, Ramlall TF, Webb WW, Eliezer D. Quantification of alpha-synuclein binding to lipid vesicles using fluorescence correlation spectroscopy. Biophys J. 2006;90:4692–4700. [PMC free article] [PubMed]
9. Chen H, Rhoades E, Butler JS, Loh SN, Webb WW. Dynamics of equilibrium structural fluctuations of apomyoglobin measured by fluorescence correlation spectroscopy. Proc Natl Acad Sci USA. 2007;104:10459–10464. [PMC free article] [PubMed]
10. Abedini A, Raleigh DP. A role for helical intermediates in amyloid formation by natively unfolded polypeptides? Phys Biol. 2009;6:015005. [PMC free article] [PubMed]
11. Crowther RA, Jakes R, Spillantini MG, Goedert M. Synthetic filaments assembled from C terminally truncated alpha-synuclein. FEBS Lett. 1998;436:309–312. [PubMed]
12. Broersen K, van den Brink D, Fraser G, Goedert M, Davletov B. Alpha-synuclein adopts an alpha-helical conformation in the presence of polyunsaturated fatty acids to hinder micelle formation. Biochemistry. 2006;45:15610–15616. [PubMed]
13. Giehm L, Oliveira CL, Christiansen G, Pedersen JS, Otzen DE. SDS-induced fibrillation of alpha-synuclein: an alternative fibrillation pathway. J Mol Biol. 2010;401:115–133. [PubMed]
14. Wakamatsu M, et al. Selective loss of nigral dopamine neurons induced by overexpression of truncated human alpha-synuclein in mice. Neurobiol Aging. 2008;29:574–585. [PubMed]
15. Michell AW, et al. The effect of truncated human alpha-synuclein (1-120) on dopaminergic cells in a transgenic mouse model of Parkinson’s disease. Cell Transplant. 2007;16:461–474. [PubMed]
16. Beyer K. Mechanistic aspects of Parkinson’s disease: alpha-synuclein and the biomembrane. Cell Biochem Biophys. 2007;47:285–299. [PubMed]
17. Li W, et al. Aggregation promoting C-terminal truncation of alpha-synuclein is a normal cellular process and is enhanced by the familial Parkinson’s disease-linked mutations. Proc Natl Acad Sci USA. 2005;102:2162–2167. [PMC free article] [PubMed]
18. Kuznetsova IM, Turoverov KK, Uversky VN. Use of the phase diagram method to analyze the protein unfolding-refolding reactions: fishing out the “invisible” intermediates. J Proteome Res. 2004;3:485–494. [PubMed]
19. Luo P, Baldwin RL. Mechanism of helix induction by trifluoroethanol: a framework for extrapolating the helix-forming properties of peptides from trifluoroethanol/water mixtures back to water. Biochemistry. 1997;36:8413–8421. [PubMed]
20. El-Agnaf OM, Jakes R, Curran MD, Wallace A. Effects of the mutations Ala30 to Pro and Ala53 to Thr on the physical and morphological properties of alpha-synuclein protein implicated in Parkinson’s disease. FEBS Lett. 1998;440:67–70. [PubMed]
21. Serpell LC, Berriman J, Jakes R, Goedert M, Crowther RA. Fiber diffraction of synthetic alpha-synuclein filaments shows amyloid-like cross-beta conformation. Proc Natl Acad Sci USA. 2000;97:4897–4902. [PMC free article] [PubMed]
22. Ahmad MF, Ramakrishna T, Raman B, Rao C. Fibrillogenic and non-fibrillogenic ensembles of SDS-bound human alpha-synuclein. J Mol Biol. 2006;364:1061–1072. [PubMed]
23. Conway KA, et al. Acceleration of oligomerization, not fibrillization, is a shared property of both alpha-synuclein mutations linked to early-onset Parkinson’s disease: implications for pathogenesis and therapy. Proc Natl Acad Sci USA. 2000;97:571–576. [PMC free article] [PubMed]
24. Pountney DL, et al. Annular alpha-synuclein species from purified multiple system atrophy inclusions. J Neurochem. 2004;90:502–512. [PubMed]
25. McClendon S, Rospigliosi CC, Eliezer D. Charge neutralization and collapse of the C-terminal tail of alpha-synuclein at low pH. Protein Sci. 2009;18:1531–1540. [PMC free article] [PubMed]
26. Chandra S, Chen X, Rizo J, Jahn R, Sudhof TC. A broken alpha -helix in folded alpha -synuclein. J Biol Chem. 2003;278:15313–15318. [PubMed]
27. Bussell R, Jr, Eliezer D. A structural and functional role for 11-mer repeats in alpha-synuclein and other exchangeable lipid binding proteins. J Mol Biol. 2003;329:763–778. [PubMed]
28. Ulmer TS, Bax A, Cole NB, Nussbaum RL. Structure and dynamics of micelle-bound human alpha-synuclein. J Biol Chem. 2005;280:9595–9603. [PubMed]
29. Georgieva ER, Ramlall TF, Borbat PP, Freed JH, Eliezer D. Membrane-bound alpha-synuclein forms an extended helix: Long-distance pulsed ESR measurements using vesicles, bicelles, and rodlike micelles. J Am Chem Soc. 2008;130:12856–12857. [PMC free article] [PubMed]
30. Drescher M, et al. Spin-label EPR on alpha-synuclein reveals differences in the membrane binding affinity of the two antiparallel helices. Chembiochem. 2008;9:2411–2416. [PubMed]
31. Bodner CR, Dobson CM, Bax A. Multiple tight phospholipid-binding modes of alpha-synuclein revealed by solution NMR spectroscopy. J Mol Biol. 2009;390:775–790. [PMC free article] [PubMed]
32. Necula M, Chirita CN, Kuret J. Rapid anionic micelle-mediated alpha-synuclein fibrillization in vitro. J Biol Chem. 2003;278:46674–46680. [PubMed]
33. Ferreon AC, Deniz AA. Alpha-synuclein multistate folding thermodynamics: implications for protein misfolding and aggregation. Biochemistry. 2007;46:4499–4509. [PubMed]
34. Li J, Uversky VN, Fink AL. Effect of familial Parkinson’s disease point mutations A30P and A53T on the structural properties, aggregation, and fibrillation of human alpha-synuclein. Biochemistry. 2001;40:11604–11613. [PubMed]
35. Fredenburg RA, et al. The impact of the E46K mutation on the properties of alpha-synuclein in its monomeric and oligomeric states. Biochemistry. 2007;46:7107–7118. [PubMed]
36. Bussell R, Jr, Eliezer D. Effects of Parkinson’s disease-linked mutations on the structure of lipid-associated alpha-synuclein. Biochemistry. 2004;43:4810–4818. [PubMed]
37. Ulmer TS, Bax A. Comparison of structure and dynamics of micelle-bound human alpha-synuclein and Parkinson disease variants. J Biol Chem. 2005;280:43179–43187. [PubMed]
38. Vilar M, et al. The fold of alpha-synuclein fibrils. Proc Natl Acad Sci USA. 2008;105:8637–8642. [PMC free article] [PubMed]
39. Bussell R, Jr, Eliezer D. Residual structure and dynamics in Parkinson’s disease-associated mutants of alpha-synuclein. J Biol Chem. 2001;276:45996–46003. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...