Ammonium hydroxide treatment of Aβ produces an aggregate free solution suitable for biophysical and cell culture characterization

Aβ HFIP and NH₄OH pretreatment

Synthetic human Aβ₄₂ peptide was prepared in two separate ways. Firstly the Aβ₄₂ lyophilised powder was dissolved in TFA and dried under a nitrogen stream. The remaining film was dissolved in 100% HFIP to a concentration of 1 mg/ml, sonicated 5 min in a bath sonicator and dried under a nitrogen stream. The HFIP treatment was repeated twice more and on final dissolving the peptide was dispensed into microcentrifuge tubes. After drying under a nitrogen stream, the peptide was further dried under vacuum for 1–2 h to give a clear film.

The second method was to take the original peptide and dissolve it in 10% (w/v) NH₄OH at 0.5 mg/ml. The peptide was incubated for 10 min at room temperature followed by sonication (5 min) and then dispensed (0.5 ml) into microfuge tubes. The NH₄OH was removed by lyophilisation to yield a salt free fluffy white peptide. All aliquots from both methods were then stored at −80°C. Immediately prior to use, the HFIP- and NH₄OH-treated and untreated Aβ₄₂ were dissolved in 60 mM NaOH and the concentration determined by absorbance at 214 and 280 nm using extinction coefficients of 76848 M^-1 cm^-1 or 1490 M^-1 cm^-1, respectively. In all cases the resuspended peptide was analysed by mass spectrometry and the peptide found to be unmodified with no trace of HFIP or NH₄OH.

Thioflavin T assay

The ThT assays were conducted as described previously (McColl et al., 2009; Ryan et al., 2012), Briefly, Aβ₄₂ reconstituted in 60 mM NaOH to a stock concentration ∼200 µM was diluted to a final concentration of 5 µM in phosphate buffered saline (PBS, 136.89 mM NaCl, 2.68 mM KCl, 6.39 mM Na₂PO₄, 1.47 mM KH₂PO₄) pH 7.4 containing 30 µM ThT. Vehicle controls of 60 mM NaOH were used to confirm that the typical dilution of between 1 in 50 to 1 in 200 did not significantly alter the pH of the PBS solution. The samples were dispensed (200 µl) into a 96 well clear bottom plate and placed in a Flexstation 3 plate reader (Molecular Devices, Sunnyvale, CA), incubated at 37°C with agitation every 7 min. The ThT fluorescence was measured every 7 min for 24 h using excitation and emission wavelengths of 440 nm and 485 nm, respectively.

Transmission Electron Microscopy (TEM)

Carbon-coated 300-mesh copper grids were glow-discharged in nitrogen to render the carbon film hydrophilic. Samples were gently agitated, before pipetting (4 µl) onto the grids. After 30 s adsorption time, the excess was drawn off using Whatman 541 filter paper. The grids were stained with 2% w/v potassium phosphotungstate, pH 7.2, for 10 s. Grids were air-dried before examination. The samples were examined using a Tecnai 12 Transmission Electron Microscope (FEI, Eindhoven, The Netherlands) at an operating voltage of 120 KV. Images were recorded using a Megaview III CCD camera and AnalySIS camera control software (Olympus).

Dynamic light scattering

For dynamic light scattering (DLS) experiments the Aβ₄₂ dissolved in 60 mM NaOH (Stock concentration 500 µM) was diluted into 50 mM Tris pH 8.0 buffer containing 10 mM EDTA to a final concentration of 100 µM. This buffer was chosen to minimise aggregation as Aβ aggregates at a slower rate at this pH and EDTA was included to prevent metal induced aggregation (Huang et al., 2004). Thus, the aggregation state of the resuspended material was not expected to change significantly during the time taken to prepare the peptide for DLS measurements. Vehicle controls were prepared at the concentrations used in the Aβ₄₂ experiments. The pH of all reactions were checked to be pH 8.0. All reaction mixtures were filtered through a 0.2 µm syringe filter before the concentration of the sample was measured to determine the degree of peptide loss during filtration. In all cases the concentration of peptide changed less than 5%. Subsequently, 20 µl of the filtrate was loaded in quadruplicate into a Greiner 384-well low volume glass bottom plate. The plate was centrifuged (5 min, 1000 × g) to pellet any remaining larger insoluble material prior to measurement. A total time of 20 min elapsed between the addition of the Aβ₄₂ to the buffer and the acquiring of data from the plate.

The DLS measurements were taken at 25°C using a Dyna Pro Nano Star plate reader at the Bio21 Collaborative Crystallisation Centre at laser wavelength 830 nm, with temperature control capacity and auto-attenuation function. To ensure accurate readings, the final distribution was taken from an accumulation of 50 individual, 5 s DLS collections. Hydrodynamic radius (R_h) was calculated from the diffusion coefficient by the Stokes–Einstein relationship.

Size Exclusion Chromatography (SEC)

For SEC measurements the Aβ₄₂ stock in 60 mM NaOH was diluted into PBS to a final concentration of 200 µM and incubated at 25°C for 20 min. The samples were centrifuged (5 min, 16,500 × g) immediately before loading onto a S200 Superdex column (GE, 10/300 GL). The column was eluted with 50 mM Tris pH 8.0 at a flow rate of 0.5 ml/min for 1.5 column volumes. The elution of proteins was measured using a flow cell absorbance detector set at 280 nm. Molecular weight was estimated by comparison to the SEC protein calibrants (GE Healthcare).

Small-angle X-ray scattering (SAXS) measurements

A slightly modified approach for solubilising the treated Aβ₄₂ was used for SAXS measurements, whereby the lyophilised peptide (either treated with HFIP or NH₄OH, and typically in 3–4 mg aliquots) was taken up in 20 µl of 60 mM NaOH, which was immediately diluted to 13 mM NaOH with distilled water. This sample was sonicated for 5 min in a bath sonicator to ensure the peptide entered the solution at the lower concentration of NaOH. The sample was then neutralised with 10X PBS, resulting in a solution with a pH of 7.4 and centrifuged at 16,500 × g in a benchtop microcentrifuge for 10 min. The 214 nm absorbance of the supernatant was measured to determine concentration using the above quoted extinction coefficient. Typically this process resulted in stock solutions of approximately 5–7 mg/ml (1300–1500 µM). No differences in the kinetics of fibril formation, as measured by ThT fluorescence, were observed for direct dilution from 60 mM NaOH or 13 mM NaOH.

SAXS measurements were acquired at the SAXS/WAXS beamline of the Australian Synchrotron (Clayton, Victoria, Australia). Aβ₄₂ solutions, made using untreated, HFIP treated and NH₄OH treated lyophilised peptide, solubilised using the above described NaOH solubilisation protocol (1 mg/ml, 221 µM) were analysed at a camera length of 3.3 m corresponding to a range of momentum transfer 0.005 ≤ q ≤ 0.35 Å⁻¹ (where q = 4πsinθ/λ, 2θ is the scattering angle and λ is the X-ray wavelength: 1.03 Å at 12 KeV), using a Pilatus 1 M camera (Dectris, Switzerland). Data were normalised using an integrated beamstop and intensities put on an absolute scale using distilled water as a standard. To limit radiation damage, 50 µl samples/buffers were flowed through a 1.5 mm quartz capillary and 10 frames of 1 s exposure collected. The individual frames were compared for agreement before being averaged with the scatterBrain IDL program, downloaded from the Australian synchrotron SAXS/WAXS beam line website (http://www.synchrotron.org.au/).

SAXS data were analysed using Primus (Konarev et al., 2003) from the ATSAS (Petoukhov et al., 2007; Petoukhov et al., 2012) software suite (www.embl-hamburg.de/biosaxs/software.html), which was used to plot the data and create Kratky plots. A Kratky plot, is a plot of Intensity (I) multiplied by the square of the scattering vector, s, (defined as equal to 4πsin(θ)/λ, where 2θ is the total scattering angle and λ is the wavelength of the X-ray) against s (Doniach, 2001).

Aβ_1-42 toxicity in cell culture

Rat pheochromocytoma (PC12) cells were maintained and passaged every 3–4 days in PC12 media (F12K media containing 15% (v/v) horse serum, 2.5% (v/v) fetal bovine serum (FBS), 100 U/ml penicillin G sodium, 100 µg/ml streptomycin sulfate and 0.25 µg/ml amphotericin B. The PC12 cells (passage 5–15) were trypsinised (0.05% (w/v) trypsin/0.48 mM EDTA in PBS) and harvested in the maintenance media. The cells were centrifuged (400 × g, 5 min) and the cell pellet was resuspended in F12K media containing 0.5% (v/v) FBS, 100 U/ml penicillin G sodium, 100 µg/ml streptomycin sulfate and 0.25 µg/ml amphotericin B). The centrifugation step was repeated and the cell pellet was resuspended in the media containing 0.5% (v/v) FBS with the addition of 100 ng/ml NGF. The cells were seeded at 5 × 10³ cells/well (100 µl/well) into the inner 60 wells of a 96-well plate coated with poly-L-lysine. The outer 36 wells were used as a moat and were filled with sterile water (100 µl/well). The cultures were maintained at 37°C with 5% CO₂ in a humidified incubator (Heraeus Hera cell 150, Kendro Laboratory Products, Langenselbold, Germany) and allowed to differentiate for 96 h before treating with Aβ₄₂.

The HFIP- or NH₄OH-treated peptide were prepared as described for the SAXS measurements, except 0.5 mg of Aβ was reconstituted using 60 mM NaOH, sterile water and 10 × Dulbecco’s PBS consisting of 1368.9 mM NaCl, 26.8 mM KCl, 63.9 mM Na₂PO₄, 14.7 mM KH₂PO₄, 9.0 mM CaCl₂ and 4.9 mM MgCl₂ (Dulbecco & Vogt, 1954) instead of the standard laboratory reagents. The resulting Aβ peptide was assessed for concentration as described for the SAXS measurements. This process resulted in Aβ stock solutions of 180–250 µM. The concentrated peptide was diluted into the cell culture media A (F12K, 0.5% FCS, 100 ng/ml NGF, 100 U/ml penicillin G sodium, 100 µg/ml streptomycin sulfate and 0.25 µg/ml amphotericin B) at no more than 1/10 so as not to affect the composition of the culture media. The vehicle was prepared in the culture media using the same dilution as the peptide. The media/vehicle was used to make subsequent dilutions of the peptide (10–20 µM). Cultures were also treated with the toxin control, staurosporine (1 µM).

The cultures were incubated and assessed for viability after 48, 72 and 96 h using the CCK-8 reagent. Five minutes prior to the assay, a second toxin control (0.1% (v/v) Triton X-100) was added to the designated cells. The viability assay was done by aspirating the treatment media and replacing with cell culture media A containing 10% of the CCK-8 reagent (100 µl/well). The cultures were incubated for 2 h at 37°C and then measured for absorbance changes on the LabSystems Multiskan MS multiplate reader (ThermoFisher Scientific, Scoresby, Australia) using a 405 nm filter. Wells containing CCK-8 reagent and no cells were used to measure the background absorbance of the CCK-8/media and the values were subtracted from the test data. The absorbance readings for the untreated cells were normalised to 100% to adjust for variation between control values of individual experiments and the test data were expressed as a proportion of the untreated cells. The normalised data represent the mean and standard error of the mean (SEM) from 2–4 experiments. The normalised comparisons and the results were taken to be significant when the p value was below 0.05. The ANOVA and comparisons were done in a Microsoft^® Office Excel spreadsheet.

Effect of solubilisation pretreatment method on Aβ amyloid formation

The effect of HFIP treatment on amyloid formation by Aβ₄₂ peptides was investigated using a continuous ThT assay. ThT fluorescence was measured for solutions containing untreated, HFIP or NH₄OH pretreated Aβ₄₂ (Fig. 1A) and resuspended using the standard NaOH protocol. The ThT fluorescence timecourse of untreated Aβ₄₂ was variable with an average t½ of approximately 14.9 ± 2.6 h (Fig. 1A). Pretreatment with HFIP decreased the t½ by 6 h to approximately 9 ± 1.9 h, while NH₄OH pretreatment increased the average t½ to approximately 16.2 ± 1.3 h. In all cases we observe an initial increase in the ThT fluorescence intensity over the first hour, equivalent to approximately 10 arbitrary units, which is within the background of the timecourse assay and does not differ with Aβ₄₂ treatment. This increase could be attributed to a number of early events, ranging from temperature equilibration in the instrument through to the formation of cross-β sheets structures and early oligomers.

Aggregation state of Aβ after dissolution.

Due to the observed increase in the rate of fibril formation that occurs with HFIP pretreatment, we hypothesised that this solvent was not completely removing the oligomers, but was in fact allowing amyloid seeds to persist through the complete removal of HFIP and thereby accelerate aggregation and fibril formation. To investigate the starting aggregation state of Aβ₄₂ in solution we used dynamic light scattering, size exclusion chromatography and small angle X-ray scattering. DLS was used to investigate qualitative differences in the preparation techniques. DLS provides an approximate hydrodynamic radius based on the ability of a solution to scatter light and is sensitive to the presence of aggregates. While hydrodynamic radius is typically applied to spherical particles, in this case the approximate radius is sufficient to indicate a qualitative difference between the two treatments. The DLS measurements on average demonstrated that both the HFIP and NH₄OH treated Aβ₄₂ exist as a monomodal polydisperse solution (Fig. 2A) however, the spread of data for both samples was quite wide, with the extremes of low and high distributions shown in the supplementary data (Fig. S1). For the HFIP preparations we found a large degree of variability in the average distribution of five separate experiments with polydispersity factors ranging from 10%–52% and radii ranging from 3–100 nm (Fig. 2A, red line). Given this variability, the most commonly observed average radius was 8 nm with approximately 50% polydispersity. In contrast, Aβ₄₂ pretreated with NH₄OH was more consistent, with the average distribution of five separate experiments showing radii ranging from 1–10 nm and 30% polydispersity factors. The average radius of this sample was 3.5 nm with 30% polydispersity (Fig. 2A).

Size exclusion analysis (Fig. 2B) of HFIP and NH₄OH pretreated Aβ₄₂ were consistent with the findings of the DLS experiments (Fig. 2A). We observed that the treatment of Aβ₄₂ with HFIP had an elution profile that displayed a heterogeneous mix of species, ranging in elution volume from 8.5 ml (the void volume of the column) to approximately 20 ml (the total volume of the column) indicating a large range of species. Using the area under the peaks to estimate the percentage of protein present, the major peaks from the HFIP pretreated peptide occur at 8.5 ml (∼65% total protein) and 17 ml (∼20% total protein) indicating the presence of monomeric (∼5,000 Da) and large oligomeric species (MW greater than 80,000 Da). For NH₄OH pretreated Aβ₄₂ we observed a different bimodal distribution with a major (∼90% total protein) monomeric peak (MW ∼5,000 Da) eluting at 17 ml and a second minor (∼10% total protein) peak eluting at 12 ml, corresponding to a medium sized oligomeric species (MW ∼40,000–50,000 Da).

Electron microscropy of the preparation techniques

Aβ prepared for the ThT and SEC assays were analysed by TEM. Micrographs show that after 6 h at 37°C the Aβ₄₂ pretreated with HFIP has more protofibril/early fibril structures compared with the Aβ₄₂ pretreated with NH₄OH (Figs. 3A and 3B). The smaller structures of the latter are beyond the sensitivity of this technique (Fig. 3B). In light of the differing buffer conditions used for DLS, the HFIP and NH₄OH pretreated Aβ₄₂ prepared for DLS were also analysed by TEM (Figs. 3C and 3D). The representative micrographs show that some short fibrils are present in the HFIP pretreated Aβ₄₂ solution (Fig. 3C), whilst only small, unidentifiable structures are present in the NH₄OH pretreated peptide solution (Fig. 3D).

Small angle X-ray scattering of HFIP or NH₄OH pretreatment

In addition to the DLS and SEC measurements, we also investigated the initial aggregation state by SAXS measurements (Fig. 4). A comparison of the SAXS data of the untreated (Fig. 4A), HFIP treated (Fig. 4B) and NH₄OH treated (Fig. 4C) Aβ₄₂ highlights the different solution behaviour of these samples. The data indicates that all of the peptide solutions contained a polydisperse mix of aggregates, as the data in the low q region of the scattering curves displays a significant increase in intensity, highlighted in Fig. 4D. This makes interpretation of the SAXS data in terms of shape or molecular weight difficult as the scattering from the aggregated material significantly distorts the analysis, however we can transform the data into a Kratky plot, which can give an indication of the folded/unfolded state of a protein (Doniach, 2001; Mertens & Svergun, 2010), with folded proteins typically showing a prominent peak in the curve, while unfolded proteins show a continuous increase as s increases (for review of the basis of Kratky plots and on the analysis of SAXS data see (Doniach, 2001; Putnam et al., 2007)). The treated and untreated Aβ₄₂ show a qualitative difference in the potential degree of folded peptide represented by the scattering curves (Fig. 4E). NH₄OH pretreated Aβ₄₂ shows a completely unfolded structure, whilst HFIP pretreated peptide appears to represent partially folded species (Fig. 4E). The untreated peptide is intermediate in the degree of folded and unfolded peptide compared to HFIP and NH₄OH treatment.

The effect of pretreatment on cellular toxicity

Pretreatment of Aβ₄₂ has a large impact on the oligomerisation state of Aβ₄₂, which is a key factor for toxicity in vitro. A CCK-8 toxicity assay was used to investigate how the Aβ₄₂ preparation method affects toxicity in PC12 cultures after 24–96 h. Cultures treated with both NH₄OH and HFIP pretreated Aβ₄₂ (15–20 µM) were viable after 48 h (Fig. 5A). Compared with vehicle treated cells, the PC12 cultures treated with NH₄OH pretreated Aβ₄₂ (15 and 20 µM) for 72 h showed a 15.7% (P < 0.001) and 30.2% (P < 0.001) decrease in viability, respectively (Fig. 5B). In contrast, PC12 cultures treated with the HFIP pretreated Aβ₄₂ (10–20 µM) for 72 h did not show any loss of viability (Fig. 5B). However, the PC12 cultures treated with the HFIP pretreated Aβ for 96 h, were 38.4% (P < 0.001) less viable than vehicle treated cells but only at the highest concentration tested (20 µM, Fig. 5C). After 96 h, the PC12 cultures treated with the NH₄OH pretreated Aβ (15 and 20 µM) showed a 25.2% (P < 0.001) and 42.1% (P < 0.001) decrease in viability, respectively compared with vehicle treated cells (Fig. 5C).