|Home | About | Journals | Submit | Contact Us | Français|
α-Synuclein is a 140 residue protein associated with Parkinson’s disease and other neurodegenerative disorders.[1, 2] The protein has three distinct regions. The N-terminal region is positively charged; the hydrophobic core [also known as the non-amyloid component (NAC)] comprises residues 61–90, and the C-terminal region is negatively charged. Its filamentous inclusions are found in dementia with Lewy bodies and multiple system atrophy. α-Synuclein aggregates formed in vitro near physiological pH are morphologically indistinguishable from those in Lewy bodies. Understanding α-synuclein’s fibrillation mechanism at the molecular level is key to discovering or designing drugs to slow, prevent, or even reverse its aggregation.
At low sodium dodecyl sulfate (SDS) concentrations (<~2 mM), α-synuclein assembles into highly ordered amyloid fibrils.[6–9] At high SDS concentrations (>~20 mM), monomeric α-synuclein forms two stable anti-parallel curved α-helices followed by an unstructured tail.[10–17] How α-synuclein switches from an unfolded monomer in buffer to a helical conformation in SDS and why SDS promotes aggregation remain unclear. Here, we identify the conformational state through which amyloid fibril formation most likely occurs by using highly sensitive 19F NMR spectroscopy and site specifically labeled a-synuclein to probe the conformational switch.
First, we used two dimensional 1H-15N Heteronuclear Single Quantum Correlation (HSQC) spectra to monitor conformational changes in 100 µM α-synuclein upon increasing the SDS concentration from 0.8 mM to 24 mM at 25 °C (Figure 1). No aggregation occurred during the 20 min NMR experiments. As observed previously,[9, 18] resonances from residues in the C-terminal region remain unchanged, but the resonances from residues in the N-terminal region are broadened beyond detection in 0.8 and 2.4 mM SDS. The line broadening could arise from chemical exchange between the free and micelle-bound states occurring at rates approximating the difference in resonance frequencies in the two environments.[8, 9] In 24 mM SDS, additional new resonances appear, but they are broad compared to the relatively sharp resonances from the C-terminal residues.
The 19F chemical shift is sensitive to local environment and it is widely used to monitor conformational transitions in proteins,[8, 19–27] DNAs and RNAs.[29, 30] We made four variants of α-synuclein by substituting the unnatural amino acid L-4-trifluoromethylphenylalanine (tfmF) at phenylalanine 4, tyrosine 39, valine 70 and tyrosine 133. These positions were chosen to provide a probe in each of the three regions descibed above. Residues 4 and 39 are present in the N-terminal region, residue 70 is in the NAC-region, and residue 133 is in the C-terminal region. We chose these particular residue types because their aromatic or bulky hydrophobic side chains are similar.
One-dimensional 19F NMR spectra were recorded in the absence and in the presence of 0.8, 2.4 and 24 mM SDS (Figure 2). The conditions and the molar ratio of protein to SDS were kept the same as those used to obtain the HSQC data.
Without SDS, the 19F resonances from tfmF labeled proteins are sharp and show little chemical shift dispersion, as expected for a disordered protein. A sharp transition (~0.5 ppm shift) for the F4tfmF resonance takes place at 0.8 mM SDS, while the transitions for the Y39tfmF (~0.6 ppm shift) and V70tfmF (~1.2 ppm shift) resonances take place at 2.4 mM SDS. Y133tfmF does not exhibit any change from 0 to 24 mM SDS. Many studies show that α-synuclein adopts α-helical structure in SDS.8–11 Our data pinpoint the location of the changes at the residue level. Such information is unavailable from methods such as circular dichrosim (CD) and fluorescence spectroscopies. Another feature of the 19F spectra is that the resonances are broad in SDS solution compared to dilute solution, except for Y133tfmF. The results from light scattering experiments show that the broadness is not due to the formation of large soluble oligomers in SDS. Instead, the broadness is caused by conformational exchange in the ensemble. These resonances sharpen when the SDS concentration is increased from 0.8 to 24 mM, indicating a dynamic interaction between α-synuclein and SDS.
Ahmad et al. found that SDS promotes α-synuclein fibril growth at an optimal concentration range of 0.5–0.75 mM. At these SDS concentrations, the α-synuclein ensemble shows enhanced amounts of hydrophobic surface. When the SDS concentration is above 2 mM, less- or non-fibrillogenic ensembles form. These ensembles are characterized by less hydrophobic surface and maximum helix content. Our data are consistent with these conclusions and others derived from isothermal titration calorimetry (ITC), CD spectroscopy, hydrophobic dye binding, and fluorescence lifetime data.[7, 31, 32] Due to the site specific labeling, our 19F data provide detailed conformational switch information unobtainable from other techniques.
Figure 3 describes the SDS-induced conformational switch. In dilute solution, α-synuclein is disordered. CD data indicate that α-synuclein acquires α-helical switch in SDS solution. [7, 9] Changes in chemical shift from 0.8 to 2.4 mM SDS, indicate that the partial helical conformation is induced in the first several residues, including F4tfmF. But the small shift change indicates that residues after Y39tfmF remain disordered. This partial helical conformation is probably responsible for the fast aggregation of α-synuclein.
Above 2.4 mM SDS, a more folded α-synuclein with a higher helix-content forms as indicated by the changes in chemical sift observed for the V70tfM protein. The C-terminal region of this species, as monitored at position 133, remains disordered. This species is known to be less aggregation prone and fully micelle bound. These conformations exchange on the ~ms time scale based on the difference in 19F chemical shift between the states. The intensity increase and lack of chemical shift change with increasing SDS concentration indicates both dynamic changes and structural homogeneity.
α-Synuclein function is suggested to be related to membrane binding.[2, 33–36] SDS micelles are often used to mimic the membrane environment. Recent data from NMR studies of α-synuclein-membrane binding suggest that the N-terminus (residues 3–25) binds membranes as a helix, while the hydrophobic non-amyloid component region remains dynamically disordered.[37, 38] This bound state is prone to intermolecular interactions that progress toward disease-associated oligomers and fibrils. Our proposed aggregation prone state is consistent with the membrane bound state. Single-molecule fluorescence resonance energy transfer experiments have also been employed to investigate SDS induced species.[10, 39] These studies suggests that α-synuclein adopts a metastable structure. Our 19F data provide direct evidence that this metastable, aggregation prone state exists in a membrane mimetic environment.
Previously, we used 19F NMR to examine the conformational change of α-synuclein by using 3 fluorotyrosine (3FY) labeling. We showed that 19F NMR is a good probe of conformational change at the level of individual residues. tfmF labeling offers at least five advantages compared to 3FY. First, assignment is not required. Second, we obtain higher sensitivity because tfmF contains three times the 19F. Third, tfmF has advantageous NMR relaxation properties. Fourth, any single ribosomally encoded amino acid can be replaced with tfmF. Fifth, 19F NMR of tfmF labeling can also be applied to study large proteins or protein complexes (to 100 kDa), which is challenging for routine solution NMR spectroscopy.[40, 41] It must be borne in mind, however, that the trifluromethyl group may affect protein properties. Protein conformational switching is a general mechanism in signal transduction. Here, we describe a powerful and easy 19F NMR method to assess switching phenomena.
The protein was labeled and expressed by using an orthogonal tRNA/tRNA synthase system. Briefly, stop codons (TAG) were incorporated in the arabinose-inducible expression vector pBAD at the sites for labeling by using site-directed mutagenesis (QuickChange, Stratagene). The mutations were confirmed by DNA sequencing. A single DH10B colony from an ampicillin/tetracycline plate containing both the appropriate pBAD and pDule-tfm-Phe vectors was picked and used to inoculate 50 mL of LB media (containing 100 mg/L ampicillin and 25 mg/L tetracycline). The culture was grown overnight at 37 °C with shaking at 250 rpm. An aliquot (2.5 mL) was added to warm arabinose autoinduction medium (500 mL). tfmF was added after 30 min (1 mM final concentration) from a stock solution (100 mM) prepared by dissolution in NaOH (20 mM). This culture was grown for an additional 40 h (absorbance at 600 nm of ~5). Cultures were centrifuged at 1 200g for 30 min at room temperature. The cell pellets were resuspended in lysis buffer (20 mL, 10 mM Tris, pH 8.0, 1 mM EDTA, 1 mM PMSF), sonicated (FisherScientific, Sonic Dismembrator Model 500) on ice for 5 min with a duty cycle of 1 s on, 2 s off. The sonicated sample was boiled for 20 min (to precipitate globular proteins) and then centrifuged at 16 000g for 30 min. Nucleic acids were precipitated with streptomycin sulfate (10 mg/mL, 4 °C, 30 min) and removed by centrifugation (16 000g, 30 min). Proteins were precipitated with (NH4)2SO4 (360 g/L, 4 °C, 30 min). The pellet was collected by centrifugation (16 000g, 30 min) and resuspended in low salt buffer (20 mL, 20 mM Tris, pH 7.7). Protein in low salt buffer was then loaded onto HiLoad 16/10 Q Sepharose High Performance column (GE Healthcare) in Tris (20 mM), pH 7.7 and eluted with in a NaCl gradient (0 – 1 M) over 160 min (α-synuclein elutes at ~300 mM). Fractions containing α-synuclein (analyzed by Coomassiestained SDS–PAGE) were concentrated with a 5K MWCO filter (Millipore, Bedford, Massachusetts), loaded onto a Supedex™ 75 column (GE Healthcare) and eluted (150 mM NaCl, 20 mM Tris, pH 8.0). Fractions containing α-synuclein were combined and lyophilized. Purity was assessed by using SDS-PAGE. The identity of the proteins was confirmed by using mass spectrometry (NanoESI-MS).
19F NMR spectra were acquired on an INOVA 600 MHz spectrometer equipped with 19F(1H) probes at 25 °C. The 19F spectra comprised 512 transients, a 30 kHz sweep width, with a 2 s delay between transients. 19F chemical shifts are referenced to trifluoroethanol at 0 ppm. All spectra were acquired under similar conditions.
We thank Ryan A. Mehl and Peter Lansbury for supplying the expression systems, Gregory B. Young and Marc ter Horst for NMR spectrometer maintenance, and Elizabeth Pielak for comments on the manuscript. This work was supported by the National Science Foundation (MCB-051647) and the National Institutes of Health (DP1OD783).