Optical signature of retinal Tau fibrillation

Light side scattering spectroscopy

Experimental setup design

The custom bench optical setup was designed to acquire the scattering spectrum of protein aggregates, utilising a minimal sample solution volume, typically within the range of a few hundred microliters. The sketch of the experimental setup is visualised in Fig. 3. The subtraction of the reflection of the source light from optical components is a critical procedure to achieve an unbiased measurement of the side scattering spectrum, especially as the scattered light intensity is much lower than the source. Thus, our experiment has been designed to incorporate a series of advantageous features, such as the minimisation of the required liquid volume and reduced effects from cell reflections.

Optical path for light source

The light source, an OSL2 Fiber Illuminator (Thorlabs, Inc.), passes through an optical path that includes a field lens (L1) and lens a preliminary collimating lens (L2). Then, the light is focused (with L3) onto a pinhole, where the light outside of the focal point is truncated for further collimation thanks to objective (obj1). The collimated beam is polarised and ultimately focused into the capillary containing the sample. This system ensures a homogeneous light spot in the sample solution with a diameter of 0.6 mm and irradiance of 3 W/m², allowing it to enter the capillary without interacting with its sidewalls. Here, the incident light interacts with the particles in the solution.

Disposable capillaries and customised holder

A customised holder for capillaries was designed and 3D printed, enabling microliter volume analysis. Moreover, the holder geometry ensures capillary self-alignment. The used Hollow Square Borosilicate Glass Capillaries ID 0.80 × 0.80 mm (CM Scientific Ryefield (EU) Ltd.) were disposed of after each experiment to prevent contamination and thus maintain measurement reproducibility and accuracy.

Spectra acquisition and focus monitoring

After the interaction of light with the sample solution, the light passes through two separated optical paths, which are selected through a removable mirror (M4): (i) back-reflected light and (ii) side-scattered light collected by objective (Obj2) perpendicular to the incoming light. At both paths, the light passes through a polariser (analyser) to further eliminate the residual specular reflected light. Back-reflected light enables monitoring of the light spot position in the capillary on camera (C1) before and during the experiment while obtaining the spectral image at the second camera (C2). The spectral image is created in the spectrometer part of the setup composed of slit (100 μm), reflective grating (Thorlabs, Inc.) and then the second camera, where is acquired the spectral image (x vertical location, y spectrum). Spectrum wavelength was calibrated with spectral measurement, where broad wavelength light source was filtered with multiple bandpass filters, enabling the mapping of C2 camera pixels with selected wavelength lines. Before the start of each measurement, a dark and white spectrum is used to normalise spectra (Eq. 1). As a white sample, a spectrum of 20 mg/ml of ZnO in a water solution was acquired. The background signal, acquired by analysing PBS (Phosphate Buffered Saline) solution, is subtracted from the sample spectra:

Design of experimental setup. The presented setup is composed of the following components: Light Source; 9 achromatic lenses with focal lengths of 25 mm (L1, L3), 100 mm (L2), 50 mm (L4, L8, L9) and 75 mm (L5, L6, L7); 2 objectives (Obj1) and (Obj2); 4 mirrors M1, M2, M3 and removable mirror M4; 3 linear polariser filter P1, P2 and P3; customised 3D printed capillary holder; Pinhole; Iris; Slit; Grating and 2 cameras (C1 and C2). All the optical components are suitable for the 400–700 nm wavelength range.

Total side scattering power calculation

The system has been calibrated by measuring the side scattering intensity of solutions containing various concentrations of Polystyrene beads (PS beads). The PS beads have been chosen for our calibration as their scattering cross section can be calculated analytically as Mie scattering. Before the measurement of scattering, the calculations were conducted using MiePlot v4.6 computer program (Supplementary Fig. S2). Afterwards, the solutions of PS beads were analysed in the build side-scattering spectroscope, and the range of concentration where a single scattering occurs was determined (Supplementary Fig. S1), as in this range, a linear growth of intensity with PS beads concentration is observed. The scattering of a single bead was calculated and used as a normalisation constant in the following calculations, where the measured intensity was attributed to the physical properties of scattering particles in the solution. The measured intensity of PS beads in the solution can be expressed with the following equation:

where \(\:{\sigma\:}_{90}^{PSbeads}\) is cross-section scattering at 90 degrees calculated by Mie theory (Supplementary fig. S2), \(\:n\) particle number concentration, \(\:V\) volume of solution participating in scattering, \(\:{I}_{exp}\) detected intensity.

A similar equation can be written for the measured intensity of any scattering particles in the solution:

where \(\:{I}_{\text{S}}^{Protein}\) is the unknown scattering intensity of measured particles and \(\:{n}_{Protein}\)is their concentration., By combining the Eqs. (2) and (3), we obtain the multiplication of \(\:{I}_{\text{S}}^{Protein}\) and \(\:{n}_{Protein}\), which we call total side scattering power \(\:\varOmega\:\) and it is defined by the following equation:

where \(\:k\) is the constant assigning the measured intensity value physical interpretation—the Ω is expressed as a total side scattering intensity of the system, so how much is scattering at the given angle per liquid volume.

Aggregation size estimation with Rayleigh scattering

During the aggregation process, the mass concentration of aggregating molecules is constant, as also the aggregates’ density; thus, its total molecules’ volume \(\:{V}_{0}\) at the initial aggregating stage is equal to the total volume in the following stage \(\:{V}_{x}\):

We approximate the shape of the aggregate as a sphere with an initial radius \(\:{r}_{0}\), so the total volume is equal to the volume of a single sphere multiplied by their initial number \(\:{N}_{0}\), and this is equal to the volume of an aggregating particle (a single particle is resulting from a set of aggregated molecules) at different aggregation times with radius \(\:{r}_{x}\) multiplied by their number \(\:{N}_{x}\):

Then, the change in particle number can be calculated by the following equation:

The number of particles at the initial aggregation stage can also be expressed as the ratio between their measured total side scattering power \(\:{\varOmega\:}_{0}\) and the calculated side-scattering of a single scattering particle at the initial stage \(\:{I}_{s\_0}\):

The general measured total side scattering power \(\:\varOmega\:\) can be expressed as the number of particle \(\:{N}_{x}\) multiplied by the side-scatteing of a single particle \(\:{I}_{s}\):

where \(\:{I}_{0}\) is the intensity of source light, \(\:\lambda\:\) is the wavelength of light, \(\:n\) is the refractive index of aggregates, R is the distance from the scattering particle and \(\:{\phi\:}_{1}\), and \(\:{\phi\:}_{2}\) are detection limit angles defined by the numerical aperture of the objective (Fig. 3—Obj2). By combining Eqs. (8) and (9), we obtain an expression for the estimation of scattering particle radius with the known initial radius \(\:{\:r}_{0\:}\):

Sample preparation

Model protein BSA fibrillation and analysis

A model protein was used as a control to validate the instrumental setup. Bovine serum albumin (BSA) is known to rapidly form aggregates when subjected to high temperatures⁵⁴. Therefore, BSA 400 µM was kept at 63 °C for from 30 min up to 6 h. Before the analysis with the instrument, the fibrillation was confirmed using thioflavin T (ThT) fluorescence, dynamic light scattering (DLS), and scanning transmission electron microscopy (STEM).

Tau protein construct design and expression

This study used the K18 domain of Tau isoform 0N4R, which contains 4 microtubule-binding repeats (MTBR) extending from residues 244 to 372. The native cysteine 291 in Tau K18 was modified to alanine to prevent the formation of disulfide bonds and enhance the purification of the construct, given that many studies confirmed that the formation of intermolecular disulfide bridges is not required for dimerisation^55,56,57. The genes coding for K18 were cloned into a pET30a vector. The plasmid was inserted in Escherichia coli cells, and recombinant protein expression was induced using isopropyl β-D-1-tiogalattopiranoside (IPTG). E. coli cultures containing the recombinant protein were purchased from Genscript Biotech Corporation^®.

Recombinant K18 extraction

To extract the protein, 20 g of collected cells were resuspended in 100 mL of lysis buffer (20 mM MES, 1 mM EDTA, 0.2 mM MgCl₂, 300 mM NaCl, protease inhibitor, and DNAse, pH 6.4). Lysis was performed at 50% amplitude for 30 min, 3 s ON and 6 s OFF on ice. The lysate was centrifuged at 10,000 rpm for 30 min at 4 °C. The supernatant was separated from the cell debris and was then heated at 80 °C for 20 min to promote the precipitation of most of the undesired E. coli proteins. After this step, the sample was centrifuged again at 10,000 rpm for 30 min at 4 °C. The pellet containing the precipitated proteins was separated from the supernatant. Finally, dialysis was performed to exchange the buffer and prepare the sample for protein purification. The sample was inserted in a dialysis tubing cellulose membrane and left in 2 L of dialysis buffer (20 mM MES, 50 mM NaCl, 1 mM EDTA, 1 mM MgCl₂, pH 6.4) overnight at 4 °C. The next day, the sample was centrifuged at 10,000 rpm for 30 min at 4 °C for the last time. The supernatant was then collected and sterilised with a 0.22 μm syringe filter.

Recombinant K18 purification

The recombinant K18 was purified via two-step chromatography. The first step was a cation exchange chromatography, using a HiPrep™ SP HP 16/10 column (Cytiva) set on an ÄKTA^® Pure (Cytiva) chromatograph. Buffer A was 20 mM MES, 1 mM EDTA, 1 mM MgCl2, 50 mM NaCl, while Buffer B was 20 mM MES, 1 mM EDTA, 1 mM MgCl₂, 1 M NaCl. The method was: 2 CV 15% Buffer B, 2 CV 30% Buffer B, 2 CV 40% Buffer B, 2 CV 45% Buffer B, 2 CV 50% Buffer B and 2 CV 100% Buffer B. The final chromatogram was examined, and the selected elution fractions were analysed by gel electrophoresis. The bands on the gel allowed for the selection of the fractions containing the protein of interest. Those fractions were selected and prepared for the second chromatographic step, size exclusion chromatography. The column used was a HiLoad™ 26/600 Superdex 75 (GE Healthcare). Buffer A was H₂O, Buffer B was PBS, and the method was: 2 CV 100% Buffer B. The final concentration of recombinant K18 was determined via UV-Vis spectrophotometry. The molar extinction coefficient ε for K18 at 270 nm is 1490 M^-1 cm^-1. The final yield was 60 mg for 20 g of bacterial paste.

Recombinant K18 fibril preparation

Recombinant K18 400 µM in PBS was reduced using tris(2-carboxyethyl)phosphine (TCEP) 4 mM for 10 min at 55 °C. Subsequently, the aggregation agent sodium tripolyphosphate 1600 µM was added to initiate K18 protein aggregation in vitro. The samples were kept at 37 °C under rotation at 150 rpm. The protein was collected at multiple time points, 4 days, 1 week, 2 weeks, 3 weeks, and 5 weeks, and analysed via dynamic light scattering (DLS) and scanning transmission electron microscopy (STEM) to confirm the presence of fibrils.

Protein fibrillation analysis

Thioflavin T (ThT) fluorescence

To monitor the fibrillation kinetics of BSA and K18, a fluorescence analysis was performed. A widely used fluorophore, ThT, was selected. The fluorophore stock was prepared by resuspending 2 mg of powder in 1 mL of PBS, and the final concentration was calculated by UV-vis measurements using ε₄₂₀ = 31,600 M^-1 cm^-1. For BSA, ThT 20 µM was added to 400 µL of BSA 400 µM kept at 63 °C. Multiple time points were analysed: 0.5, 1, 2, 3, 4, 5, 20, and 24 h. One sample was left at RT and used as a control. For K18, ThT 20 µM was added to 400 µL of K18 400 µM fibrillated, as shown above. Multiple time points were analysed: 1, 2, 3, 4, 7, and 30 days. One sample was left at RT and used as a control. Fluorescence was measured using excitation at 420 nm and emission at 485 nm. All the measurements were made by using the RF-6000 fluorimeter (Shimadzu RF-6000).

Dynamic light scattering (DLS)

DLS measurements were performed using a logarithmic correlator in combination with a homemade optical set-up⁵⁸. The monochromatic and polarised beam of a He–Ne (λ = 632.8 nm) 10 mW laser is focused on the sample cuvette placed in a cylindrical glass cell filled with water for index matching and temperature control. The scattered light is focused in the core of a single-mode optical fibre that permits to obtain a very high coherence factor. The number of detected photons, measured from a Perkin Elmer photomultiplier used in single photon counting mode, is transferred to a computer through an input-output digital card from National Instruments^®. The intensity autocorrelation functions is directly computed using a software correlator as g₂(q, t) = 〈I (q, t)I (q, 0)〉/〈I (q, 0)〉², where q is the modulus of the scattering vector defined as q = (4πn/λ)sin(θ/2) (where n is the refractive index of the solvent and θ is the scattering angle, θ = 90° in the present experiment). The raw measurements, as directly obtained, without any data corrections, are shown in Fig. 2c for K18 samples before fibrillation and at different times after fibrillation (coloured symbols as described in the legend). Quantitative analysis of the measurements was obtained through a fit of the data with a conventional Kohlrausch–Williams–Watts expression:

where b is the coherence factor, τ is the relaxation time related to the motion of the particles and β the shape parameters connected to the sample polydispersity. In the case of dilute samples, as measured in the present manuscript, the radius of the particles can be calculated according to the Stokes–Einstein relation as R = K_BT/6πηD, where D is the translational diffusion coefficient calculated from τ through the relation τ = 1/q²D. Measurements have been performed at T = 20 °C K for BSA and T = 25 °C K for K18 samples while sample viscosity η has been approximated with the solvent one⁵⁹.

Scanning transmission electron microscopy (STEM)

To characterise the fibrils, STEM images of BSA and K18 were produced and analysed. BSA 100 µM was prepared as shown above. Time points 0.5 and 5 h were selected, plus the control was not subjected to high temperature. K18 100 µM was also prepared as described previously. The selected time points were 1 week and 4 weeks, plus a control not put at 37 °C. The samples were prepared for STEM imaging. The samples were diluted with H₂O from 100 to 50 µM. One drop of sample was loaded onto a carbon-coated 3 mm copper grid. The sample was left to adsorb for 3 min. The grid was washed with one drop of water for 3 min. Then, one drop of the staining solution uranyl acetate 3% was added and left for 3 min. The grid was washed twice with H₂O for 3 min. All samples were visualised using a Zeiss Auriga microscope used in STEM mode.