|Home | About | Journals | Submit | Contact Us | Français|
α-Synuclein (α-syn), a presynaptic protein implicated in Parkinson’s disease, binds copper(II) ion (1:1) with submicromolar affinity in vitro. Insights on the molecular details of soluble-and fibrillar-Cu-α-syn are gained through X-ray absorption spectroscopy. Our results indicate that the copper coordination environment (3-to-4 N/O ligands, average Cu-ligand distance ~1.96 Å) exhibits little structural rearrangement upon amyloid formation in spite of the overall polypeptide conformational change from disordered-to-β-sheet. Interestingly, we find that some population of CuII-α-syn reduces to CuI-α-syn in the absence of O2. This autoreduction event appears diminished in the presence of O2 suggestive of preceding CuI/O2 chemistry. Evidence for generation of reactive oxygen species is obtained by the observation of new emission features attributed to dityrosine crosslinks in fibrillar samples.
Amyloid formation of the presynaptic protein α-synuclein (α-syn), a process that has been implicated in the pathogenic mechanism of Parkinson’s disease,1 has been shown to accelerate in the presence of copper.2,3 In fact, Cu/protein interactions also are associated with other neurodegenerative ailments such as Alzheimer’s and prion diseases.4–9 Cu is a redox-active metal ion that can activate O2 leading to the generation of reactive oxygen species (ROS) such as O2−•, HO2−•, HO−•, and H2O2, which can in turn lead to enhanced oxidative stress and/or protein damage. In addition, metal coordination can perturb protein structure or initiate protein misfolding events, both of which could lead to altered function. In this work, the CuII coordination site of α-syn (Figure 1) is explored in the soluble and fibrillar states and evidence of CuI/O2 chemistry is reported.
α-Syn fibrils were formed by incubating equimolar CuIISO4 and α-syn (70 – 180 μM) in pH 7.0 buffer (20 mM MOPS, 100 mM NaCl) at 37 °C with agitation (450 rpm). The aged samples were then analyzed after one week. Amyloid formation (Figure 2) was confirmed by (1) the characteristic secondary structural change from unfolded to β-sheet monitored by circular dichroism (CD) spectroscopy, (2) a thioflavin T (ThT) fluorescence assay with subsequent quantum yield increase in the presence of α-syn fibrils,10 and (3) transmission electron microscopy (TEM). In all instances (1–3), the metal chelator ethylenediaminetetraacetic acid (EDTA) was added to the aged samples to probe changes in the fibril morphology following Cu extraction; however, no apparent differences were detected.11
Extraction of the metal was confirmed by employing the Trp-containing variant, F4W. Based on fluorescence quenching experiments, F4W is the most responsive CuII reporter amongst a series of site-specific Trp substitutions (Figure 1).12 Coordination of CuII to F4W results in full quenching of the Trp emission. Metal extraction can then be accomplished by addition of EDTA, which restores Trp emission in both soluble and fibrillar forms.11 Our result also suggests that the Cu site in fibrils is solvent-accessible (outside of the amyloid core), consistent with previously proposed fibril models where the amyloid core spans residues 32–100.13–15
To examine local conformational changes in CuII-α-syn upon fibril formation, we measured the respective Cu K-edge X-ray absorption (XAS) spectra. Notably, the Cu K-edge spectrum of the soluble Cu-bound α-syn (150 – 180 μM) exhibited an estimated 20% metal reduction (based on a four-coordinate CuI model) with the appearance of a characteristic CuI absorption feature at lower energy (~8983 eV) (Figure 3A).11 In contrast, after aging and forming fibrils, the edge absorption spectrum showed that the bound Cu ions are more oxidized than the soluble sample indicating that metal oxidation occurred during aggregation. We ensured that this result is reproducible and not attributable to photoreduction.11
The direct observation of CuI is in line with our previous study where we found enhanced CuII binding to F4W in the presence of dioxygen.16 As a result, we hypothesized that the differences in metal-protein affinity may be due to autoreduction (CuII → CuI) that is initiated by Met oxidation (Met → Met-O). The primary CuII binding motif of α-syn is at the N-terminus (MDVFMK) containing two nearby Met residues, which are highly susceptible to oxidation.17,18 Furthermore, the reduction of CuII → CuI has been suggested for Aβ, the amyloidogenic peptide implicated in Alzheimer’s disease.4,9 Accordingly, a similar electron transfer pathway is feasible for CuII/I-α-syn. For both systems, Aβ and α-syn, further work is necessary to confirm this claim since many possible electron donors are present in aqueous buffer solution, including water.
To assess the role of O2-chemistry, the Cu K-edge measurements were conducted on samples prepared under anaerobic conditions. We find higher percentages of metal reduction (CuII → CuI) for both soluble and aggregated samples. Similar experiments also were conducted under 100% O2 and a comparable extent of metal reduction was observed as under aerobic conditions (air ~ 21% O2).11 Taken together, our results indicate that metal reduction of CuII → CuI occurs in the absence of dioxygen and in the presence of O2 re-oxidation follows (CuI → CuII) suggesting generation of ROS.
Evidence of ROS formation and presumably CuI/O2 reactivity was gained by the observation of a new emission feature at 395 nm following fibril formation, which we attribute to the formation of dityrosines (Figure 3C).19 Interestingly, protein samples aged in the absence of Cu exhibited a less pronounced and significantly red-shifted emission at 425 nm (Figure 3C), which we also associate with the presence of dityrosine chromophores.20,21 Importantly, neither of these emission bands were detected from samples aged anaerobically. Therefore, dityrosine formation occurs in the presence of O2 and is greatly enhanced by coordinated Cu. Furthermore, the 30 nm spectral shift may indicate the involvement of different Tyr residues and possibly differences between inter- versus intra-molecular crosslinks (Figure 1).
Surprisingly, the extended X-ray absorption fine structure (EXAFS) data (Figure 3B) show only modest differences between soluble versus aggregated CuII-α-syn despite the clear differences in secondary structure content as indicated by our CD data (Figure 2A). Moreover, this result was unexpected because of the intrinsically disordered nature of α-syn; it is reasonable to anticipate a perturbation to the Cu coordination during global polypeptide rearrangement, which is necessary for amyloid formation. However, our data support coordination of three-to-four N- or O-containing ligands in both protein structures with approximate bond distances of 1.96 Å.
A tetra-coordinate CuII binding site is consistent with that proposed by Fernández and coworkers as well as our own group.17,22 In previous work, we identified the first four N-terminal residues, MDV(F/W), as the minimal CuII binding site based on Trp fluorescence measurements on synthetic peptide models.17 The free N-terminal nitrogen (NH2) is the anchoring residue and the remainder of the coordination site is thought to consist of two deprotonated backbone amides (N−) and either the Asp-2 carboxylate or an exogenous H2O molecule. Notably, we recently have reported an intense minimum observed by CD spectroscopy that points to the possible existence of a CuII-(Phe/Trp) cation-π interaction in α-syn based on our synthetic N-terminal peptide models;16 further work is underway.
For the first time, we show that upon coordination of CuII to α-syn, metal reduction from CuII → CuI occurs in the absence of dioxygen. In the presence of O2, re-oxidation of CuI → CuII takes place through the generation of ROS and over time results in dityrosine crosslinking. The XAS data indicate that the CuII coordination site of α-syn exhibits little change after amyloid formation and TEM analyses further show that the fibrils remain intact with removal of Cu. Overall, the results underscore that coordination of Cu to α-syn promotes oxidative stress, a factor largely associated with age-related diseases.2,4,23
Supported by the Intramural Research Program at NIH, NHLBI (JCL) and Cornell University (SD). XAS was carried out at SSRL, a DOE, BES facility. The SSRL SMB Program is supported by the DOE, BER and NIH, NCRR. We also thank Duck-Yeon Lee (NHLBI Protein Analysis Facility) and Grzegorz Piszczek (NHLBI Biophysical Facility).