Ligand Profiling to Characterize Different Polymorphic Forms of α-Synuclein Aggregates

The presence of amyloid fibrils is a characteristic feature of many diseases, most famously neurodegenerative disease. The supramolecular structure of these fibrils appears to be disease-specific. Identifying the unique morphologies of amyloid fibrils could, therefore, form the basis of a diagnostic tool. Here we report a method to characterize the morphology of α-synuclein (αSyn) fibrils based on profiling multiple different ligand binding sites that are present on the surfaces of fibrils. By employing various competition binding assays, seven different types of binding sites were identified on four different morphologies of αSyn fibrils. Similar binding sites on different fibrils were shown to bind ligands with significantly different affinities. We combined this information to construct individual profiles for different αSyn fibrils based on the distribution of binding sites and ligand interactions. These results demonstrate that ligand-based profiling can be used as an analytical method to characterize fibril morphologies with operationally simple fluorescence binding assays.


Materials and Instrumentation
All solvents and chemicals were obtained from commercial sources and used without further purification unless otherwise stated.Reactions were monitored by TLC or LCMS.TLC analyses were performed on Merck TLC Silica gel 60 F254 glass plates (0.2 mm).LCMS analyses of samples were performed using a Waters Acquity H-class UPLC coupled with a single quadrupole Waters SQD2.An Acquity UPLC CSH C18 Column, 130Å, 1.7 μm, 2.1 mm x 50 mm was used as the UPLC column.
Purification of compounds by silica column chromatography were performed using an automated system (Combiflash® Rf+ or Combiflash® Rf+ Lumen) with prepackaged silica cartridges (25 μm or 50 μm PuriFlash® columns). 1H and 13 C NMR spectra were recorded using a Bruker 600 MHz Avance 600 BBI spectrometer, a 500 MHZ Acance III Smart Probe spectrometer, or a 400 MHz Avance III HD Smart Probe spectrometer at 298.0 ± 0.1 K. Residual solvent peaks were used as an internal standard for calibration.All chemical shifts are quoted in ppm on the δ scale and the coupling constants are expressed in Hz.Signal splitting patterns are described as a singlet (s), broad singlet (br s), doublet (d), triplet (t), quartet (q), or multiplet (m).
HPLC-MS and HPLC-MS/MS analysis was performed on an Agilent 1100 Series LC system equipped with a G1310A isocratic pump, G1314A variable wavelength detector, G1316A thermostatted column compartment, and an Agilent 6300 Series Ion Trap.UV-vis spectra were collected on an Agilent Cary 60 UV-vis spectrophotometer controlled by Cary WinUV software.
Fluorescence spectroscopic data were recorded using an Agilent Cary Eclipse Fluorescence Spectrophotometer controlled by Cary WinUV software, and equipped with a Cary Eclipse Automated Polarizer for anisotropy measurements.FT-IR spectra were collected with an ALPHA FT-IR Spectrometer from Bruker.Melting points were recorded with a Mettler Toledo MP90 melting point apparatus.
Protein LoBind (Eppendorf) microtubes were used for preparing and storing all solutions containing protein.Low retention pipette tips were used for all aqueous fluid handling.Thioflavin T (1.13 g, 3.54 mmol) was purchased from Sigma Aldrich with dye content ≥65%.
The formed precipitate was collected using vacuum filtration and washed with cold water (200 mL) then dried in vacuo to afford S4 as a brown solid (17.4 g, 112 mmol, 100%).

Fluorescence Characterisation
Fluorescence spectral readings were performed on an Agilent Cary Eclipse Fluorescence Spectrophotometer using a scan rate of 600 nm/min, a data interval of 1.0 nm and an averaging time of 0.10 at 25°C.Fluorescence experiments used 20 nm excitation and emission slits and medium PMT voltage, except when specified otherwise.Fluorescence spectra of BTA were recorded with 10 nm excitation and emission slits, and low PMT voltage.
Fluorescence anisotropy experiments were performed using 10 nm excitation and emission slits, a scan rate of 120 nm/min, a data interval of 1.0 nm, an averaging time of 0.

UV-Visible Characterisation
Stock solutions of ligand in DMSO (10 mM) were diluted into ethanol to obtain a 50 µM solution and placed into a quartz fluorescence cuvette (Hellma Analytics) with a 1 cm pathlength.UV-visible spectra were obtained with an Agilent Cary 60 UV-vis spectrophotometer controlled by Cary WinUV software using a scan rate of 600 nm/min, a data interval of 1.0 nm and an averaging time of 0.10 at 25°C.

Preparation of αSyn Fibrils
Wild-type human monomeric αSyn in 1xPBS (180 µM) expressed in E. Coli and purified as previously reported. 10The four different morphologies were prepared based on previously reported procedures. 7,8,9ree buffers were prepared for aggregating αSyn fibrils.The conditions used were based on literature procedures that appeared to generate fibrils with different morphologies.A solution of monomeric αSyn in 1xPBS (180 µM) was added to an Amicon Ultra-15 Centrifugal filter (15 kDa MWCO) and centrifuged (15 min, 4000 x g).The retained monomeric αSyn was washed with one of the three buffers to be used for aggregation by adding 5 mL of buffer and centrifuging (15 min, 4000 x g).This wash step was repeated four times in total.The retained filtrate was diluted to 1 mL using the desired buffer and incubated at 37 °C for 72 h with gentle agitation by a magnetic stir bar in an Eppendorf LoBind microcentrifuge tube (2.0 mL).The resultant fibrils were then pelletised in a centrifuge (15 min, 4000 x g), the supernatant removed, and the fibrils gently resuspended in the desired buffer.The absorbance of monomer in the removed supernatant was measured at 280 nm (ε = 5,960 M -1 cm -1 ) to determine the concentration of fibrils, given as the concentrated of aggregated The αSyn 1s fibrils were prepared by sonicating a solution of αSyn 1 (183 µM) in 1xPBS between 20 and 30 seconds using a probe sonicator (Bandelin, Sonopuls HD 2070), using 10% maximum power and 50% cycles three times.The sample was then separated into aliquots and stored at -21 °C for 6 months until required.

Biophysical Characterisation of Amyloid Fibrils
Circular Dichroism Spectra Circular dichroism (CD) spectra of αSyn fibrils (1.0 μM) in 1xPBS (pH 7.4) were recorded with a Chirascan CD1 Spectrometer (Applied Photonics Ltd.) equipped with a Series 800 Temperature Controller (Alpha Omega Instruments).Far-ultraviolet measurements (190-250 nm) were recorded at 25 °C with a 1.0 cm optical pathlength, a time-per-point of 1.0 s, a 1.0 nm bandwidth, and a wavelength step of 0.1 nm.CD spectra were averaged over six scans.Data were baseline corrected by subtracting the complete buffer spectrum of 1xPBS (pH 7.4) averaged over six scans.Applied Photophysics Pro-Data Chirascan software was used to smooth the data using Savitsky-Golay smoothing and a window size of eight, and the data was converted to molar ellipticity.

Transmission Electron Microscopy
Nanoscale morphologies of fibril samples were observed by transmission electron microscopy (TEM) using a Thermo Scientific (FEI Company) Talos F200X G2 microscope operating at 200 kV.Images were recorded with a Ceta 4k x 4k CMOS camera.For sample preparation, TEM grids (continuous carbon film on 300 mesh Cu) were glow discharged using a Quorom Technologies GloQube at 25 mA for 60 s.A 2 µL sample of fibril in 1xPBS (1.0 µM) was placed on a freshly glow-discharged grid, and after 1.0 min was carefully removed by blotting with filter paper.The sample was negatively stained using 2.0 µL of 2% (w/v) uranyl acetate solution in ethanol for 30 s.The grid was blotted and dried in air for 10 min at room temperature before use.

In Vitro Binding Assays
General Procedure for Fluorescence Titrations Fluorescence spectra were measured using the general procedure described above for fluorescence characterisation.Stock solutions of ligand in DMSO at concentrations of 1.0 mM were prepared.
Stock solutions of αSyn fibrils in 1xPBS (pH 7.4, 10 μM) were prepared.Titration solutions were prepared by diluting ligands (1.0 mM in DMSO) and αSyn fibrils (10 μM in 1xPBS) in 1xPBS (pH 7.4) to the desired concentration.All titrations were performed in 1xPBS (pH 7.4) at 25°C.All titrations were performed alongside a corresponding dilution series in the absence of any amyloid fibrils.Titrations were repeated with at least three independent replicates.Each replicate was performed using titration solutions freshly prepared from different stock solutions.Spectral experiments detecting ThT used λex = 440 nm, and measured emissions from λem= 470 -600 nm.

Data fitting
Fluorescence spectra were analysed using a Microsoft Excel spreadsheet prepared by Professor Christopher Hunter.
The Microsoft Excel spreadsheet fitted the measured fluorescence intensity at a fixed wavelength to a 1:1 binding isotherm using purpose-written VBA macros employing two algorithms, COGS and Simplex.This spreadsheet is generalisable to multiple ligands and binding sites.For a ligand  and binding site , the intensity of the fluorescence emission () is given by Equation 1, Eq. 1 where  ! ! and  "  " are the product of the UV-vis absorption extinction coefficient and the fluorescence quantum yield for free and bound  respectively, [L] is the concentration of free , and is the concentration of  bound to S. The quantity  #,! Φ #,! was measured using dilution series in the absence of host.
Equation 2 is used to fit anisotropy data, Eq. 2 where  is the measured anisotropy of the system,  ! is the anisotropy of the free ligand, and  " is the anisotropy of the bound ligand.
The concentration of  bound to  is then given by Equation 3, where [S] is the concentration of unbound site , and  + is the dissociation constant of  binding to .The total concentration of , [L ,-, ], is then given by Equation 4, Eq. 4 The spreadsheet is generalisable to multiple ligands and binding sites, which is required for competition binding assays.The total concentration of binding site, [L ,-, ], is also optimised using this method to avoid assumptions about the stoichiometry of binding sites to protein concentration.For competition binding assays the reporting ligand (ThT) was assumed to bind to two binding sites with an identical dissociation constant to form complexes with an identical optical brightness.For two-step competition binding assays, the fitting procedure assumed that no competition occurred between the first and second competing ligands.This assumption was justified by the fact that the second competing ligand would preferentially displace ThT, which in all instances had the weaker dissociation constant.

Dilution Series
Dilution series of ligands were performed according to the general methods described.

One-Step Blocked Binding Assays
Binding assays of ThT were performed in the presence of ThR, OXI, and S5H according to the general methods.The presence of these competing ligands influences what sites the binding assays report on.
For example, on αSyn 1 fibrils, OXI occupies sites B and C. Therefore, ThT will preferentially bind to the unoccupied site A in the presence of the nanomolar ligand OXI.
The calculated ThT binding constants were all very similar (-log(Kd/M) = 5.5-6.1).However, the binding of ThT to αSyn 3 was relatively strong in the absence of any other ligands (log(Kd/M) = 6.7).
This titration reported on the binding of ThT to sites A, D, E, and F. Titrations in the presence of OXI and S5H show weaker binding and reported on sites A and D, and D and E respectively.ThT therefore has a higher than expected affinity for Site F on αSyn 3.

Tables of Quantitative Binding Measurements
Tables showing the binding constants for different ligands to different fibrils measured by each binding assay performed.The proposed binding sites targeted by each assay are shown.
Table S3.Dissociation constants for binding of ThT to αSyn fibrils.a a "n.d." indicates that the dissociation constant could not be determined from the titration data."-" indicates that the experiment did not report on this subset of sites.Boxes highlighted in blue indicate outliers, and average values of -log(Kd/M) exclude these outliers.Errors represent a 99% confidence interval calculated from at least three independent experiments.
5 s, and medium PMT voltage at 25 °C.A reference solution of BTA in DMSO (2 µM) was used to calculate G-factors.A G-Factor voltage of 425 V was used, and a polarisation/anisotropy voltage of 740 V was used.Anisotropy experiments detecting BTA used λex = 360 nm, and measured λem= 433 -453 nm.

Figure S33 .
Figure S33.Titration of ThT into OXI (1.42 µM) in 1xPBS (pH 7.4, 25 °C).Spectra were recorded using λex = 440 nm and monitoring emission at λem = 483 nm.Spectra were recorded using λex = 440 nm and emission intensity was averaged over λem = 481-485 nm.The experimental measurements are shown as points (at least three independent experiments were performed; errors bars are omitted for clarity), and the lines are the best fit to a 1:1 binding isotherm averaged over each independent experiments (αSyn 1: green circles, αSyn 1s: orange squares, αSyn 2: blue crosses, αSyn 3: black triangles).

Figure S35 .
Figure S35.Titration of ThT into ThR (5.0 µM) in 1xPBS (pH 7.4, 25 °C).Spectra were recorded using λex = 440 nm and monitoring emission at λem = 483 nm.Spectra were recorded using λex = 440 nm and emission intensity was averaged over λem = 481-485 nm.The experimental measurements are shown as points (at least three independent experiments were performed; errors bars are omitted for clarity), and the lines are the best fit to a 1:1 binding isotherm averaged over each independent experiments (αSyn 1: green circles, αSyn 1s: orange squares, αSyn 2: blue crosses, αSyn 3: black triangles).Binding to αSyn 1s and αSyn 2 was too weak to fit.
Dissociation constants for binding of BTA to αSyn fibrils.a a "n.d." indicates that the dissociation constant could not be determined from the titration data."-" indicates that the experiment did not report on this subset of sites.Boxes highlighted in blue indicate outliers, and average values of -log(Kd/M) exclude these outliers.Errors represent a 99% confidence interval calculated from at least three independent experiments.

6 -
TableS5.Dissociation constants for binding of OXI to αSyn fibrils.a a "n.d." indicates that the dissociation constant could not be determined from the titration data."-" indicates that the experiment did not report on this subset of sites.Boxes highlighted in blue indicate outliers, and average values of -log(Kd/M) exclude these outliers.Errors represent a 99% confidence interval calculated from at least three independent experiments.
that the experiment did not report on this subset of sites.Boxes highlighted in blue indicate outliers, and average values of -log(Kd/M) exclude these outliers.Errors represent a 99% confidence interval calculated from at least three independent experiments.
Dissociation constants for binding of ThR to αSyn fibrils.a a "n.d." indicates that the dissociation constant could not be determined from the titration data."-" indicates that the experiment did not report on this subset of sites.Boxes highlighted in blue indicate outliers, and average values of -log(Kd/M) exclude these outliers.Errors represent a 99% confidence interval calculated from at least three independent experiments.

Table S2 .
Emission and excitation maxima for ThT and BTA in 1xPBS or EtOH at 298 K.