Sedimentation and laser light scattering methods for quantifying synthetic tau aggregation propensity

Abstract

Tau aggregation assays detect and quantify the conversion of soluble tau monomers into species having filamentous or oligomeric structure. Assays for filamentous aggregates in cross-β-sheet conformation leverage optical, biochemical or biophysical methods, each with their own advantages and throughput capacity. Here we provide protocols for two medium-throughput assays based on sedimentation and laser light scattering, and compare their performance, their utility for characterizing tau aggregation dynamics, and their limitations relative to other approaches. Additionally, a protocol for transmission electron microscopy analysis is updated so as to be compatible with the truncated tau variants that have emerged as powerful tools for interrogating the structural basis of tau polymorphism. Together these methods contribute to a rich tool kit for interrogating tau aggregation kinetics and propensity over a wide range of experimental conditions.

1. Introduction

Tauopathic neurodegenerative diseases convert intrinsically disordered tau protein monomers into aggregates having cross-β-sheet structure and thread-like morphology. Cryo-electron microscopy analysis of tauopathy-derived aggregates reveals protofilamentous and filamentous forms that differ in the folding pattern adopted by the microtubule binding repeat region (MTBR) of tau protomers in each disease (reviewed in (1)). This variability is termed “polymorphism”, and each distinct protofilament structural variant is termed a “conformer” or “polymorph”. Currently nine polymorphs in seven tauopathies are recognized, establishing a preliminary taxonomy of tauopathic disease based solely on protein structure (2). The tau field seeks to clarify the mechanisms through which polymorphs form and how the process is impacted by post-translational modifications, interacting substances (including proteinaceous and non-proteinaceous species), and small-molecule inhibitors of potential therapeutic interest.

Tau aggregation assays performed in vitro are central to this effort. Because all six human central nervous system tau isoforms are highly soluble in aqueous solution, their aggregation over experimentally tractable time periods typically is induced by addition of exogenous anionic co-factors such as linear polymers such as heparin (3–5) or RNA (6), anionic surfaces such as those presented by micelles, vesicles and microspheres (7,8), and small-molecule anionic dyes (9,10). Aggregation also can be initiated by vigorous agitation in the presence of neutral surfaces (9,10). In contrast, spontaneous aggregation can be fostered through truncation of full-length tau isoforms so as to minimize charged amino acid residues lying outside the MTBR (11,12). Once initiated, tau fibrillation follows nucleation-elongation kinetics featuring lag, growth and plateau phases (Fig. 1). Lag and growth phases provide information on aggregation mechanisms including nucleation rate (13), whereas plateau phase can be used to estimate the minimal concentration of tau needed to support fibrillation (14,15). Together these parameters quantify the propensity of monomeric tau proteins to convert into filamentous forms.

Fig. 1.

Graphical depiction of parameters associated with tau aggregation dynamics. (a) The time course of tau aggregation at constant temperature is characterized by lag, growth and plateau phases. Parameters useful for assessing mechanism include t₅₀, the time to maximal growth rate and half reaction plateau, and the length or mass of filaments at reaction plateau. Extrapolation from t₅₀ to the abscissa intercept (dashed line) yields an estimate of lag time, which provides information on nucleation mechanism and rate. (b) Dependence of plateau aggregation on bulk tau concentration. Extrapolation of this linear relationship to the abscissa yields an estimate of the minimal concentration of tau required to support filament formation. These parameters can be extracted from even semi-quantitative approaches such as thioflavin dye fluorescence and laser light scattering.

A variety of aggregation assays are available to estimate these parameters under near-physiological conditions of tau concentration (16), pH and ionic strength, with each having its own advantages and disadvantages for detecting the aggregated state (Table 1). For example, Thioflavin dye assays, which rely on changes in the quality and/or quantity of probe fluorescence when bound to cross-β-sheet structure (17), support real-time detection of cross-β-sheet aggregate formation in solution (18,19). The approach is especially powerful because it does not require separation of bound from free probe or aggregated from non-aggregated tau, and because it is compatible with high-throughput plate readers. The approach is limited, however, by being incompatible with compounds that absorb light or autofluoresce at visible wavelengths, including certain small-molecule aggregation inhibitors (20). Filter trap assays also represent a high throughput option owing to the ability of semipermeable membranes to separate aggregation products from tau monomer reactants in multi-well format, while being fully compatible with optically active small molecules. They also offer unparalleled sensitivity and specificity when combined with tau immunodetection, and so can be effective with protein mixtures. However, the method requires careful calibration with standards owing to possible detection nonlinearity observed with anti-tau antibodies (21). Transmission electron microscopy (TEM) assays leverage adsorption of tau fibrils to grids (9) and subsequent imaging after negative staining. The TEM approach is compatible with optically active compounds while providing definitive description of filament morphology and length at ~20 Å resolution (22). Moreover, the combination of total filament length and length distribution provides the most rigorous data source for deducing aggregation kinetic mechanism (9,10). Tau aggregate ultrastructure also can be captured by atomic force microscopy, which shares with TEM the ability to capture tau filament length distribution (23). Nonetheless, both approaches are limited to low-throughput analysis while requiring substantial expenditures on equipment and reagents.

Table 1.

Assays for monitoring tau aggregation in vitro. The strengths of each method are listed along with considerations in selecting an appropriate assay.

Method	Application examples	Strengths	Limitations	Detailed protocols
Thioflavin dye binding	Real time, in-solution aggregation dynamics (49) Aggregation inhibitor characterization (18,19,50) Ligand displacement assays (50)	High throughput Compatible with all stages of aggregation kinetics (lag, elongation, and plateau)	Incompatible with compounds that absorb or auto-fluoresce within dye excitation/detection range	(24,51)
Filter trap	Small-molecule aggregation inhibitors (20,52)	High throughput High sensitivity	Immunodetection can be nonlinear Not specific for filamentous vs. amorphous aggregates	(21,24)
Transmission Electron Microscopy	Filament length distributions (53) Aggregation dynamics (54,55) Direct visualization of filament morphology (54,55) Small-molecule aggregation inhibitors (20)	Compatible with all inducers and inhibitors Sensitive and specific for fibrillation	Low throughput Expensive	(9,24)
Sedimentation	Quantification of aggregation in plateau phase (18,19)	Medium throughput Compatible with most inducers, inhibitors etc.	Inappropriate for interrogating aggregation dynamics Not specific for filamentous vs. amorphous aggregates	Herein
Laser light scattering (LLS)	Aggregation dynamics (56) Aggregation inhibitor assays (57) Quantification of aggregation in plateau phase (27,57)	Medium throughput Compatible with all stages of aggregation kinetics (lag, elongation, and plateau)	Data affected by optically active compounds and aggregates Not specific for filamentous vs. amorphous aggregates	(27), Herein

Previously we described protocols for three of these assays (Thioflavin dye fluorescence, filter trap and TEM; (24)). Here we describe two additional medium-throughput methods for quantifying tau aggregation (Table 1). The first is the sedimentation assay, which uses centrifugal force to separate aggregates from free tau proteins. Products of the aggregation reaction partitioning between supernatant and pellet fractions are then visualized by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) followed by simple Coomassie blue staining or immunoblot detection with anti-tau antibodies. The assay allows quantification of tau aggregation at plateau in the presence of a wide range of inducers, interacting proteins and optically active molecules and is capable of doing so under conditions that meet classic definitions of protein solubility (100,000×g × 1 h; reviewed in (25)). However, it is not suitable for analysis of tau aggregation dynamics and cannot distinguish fibrillar from amorphous aggregation products.

The second assay leverages laser light scattering (LLS) by tau aggregates to quantify their presence without need for a detection probe or for separation of aggregate products from reactant tau (26,27). The approach also is non-destructive and can be used to quantify aggregation dynamics. However, like Thioflavin dye assays, optically active compounds can interfere with this method depending on their light absorbance and scattering properties.

Finally, we update our TEM protocol to make it compatible with truncated tau constructs used increasingly in aggregation assays. These tend to bind carbon-coated TEM grids with weaker affinity relative to full-length tau. The update has been validated with recombinant 2N4R A2_L243del T373_L441del (K18 construct; (28)) as a representative tau fragment corresponding to the MTBR.

2. Materials

2.1. Tau aggregation reagents and supplies

1. Recombinant full-length and/or mutant tau proteins: these are prepared by liquid chromatography after E. coli expression using established protocols (see Note 1).

2. Tau aggregation inducer: choice depends on application (see Note 2).

3. 5x Aggregation Buffer: 500 mM NaCl, 50 mM HEPES pH 7.4, 5 mM DTT (see Note 3).

4. Low protein binding pipette tips and tubes (see Note 4).

2.2. Sedimentation Assay

1. Refrigerated ultracentrifuge, rotor and tubes compatible with 100,000×g forces (see Note 5).

2. Apparatus for SDS-PAGE and associated reagents (see Note 6).

3. ImageJ software (see Note 7).

2.3. Light Scattering Assay

1. Laser light source (see Note 8).

2. Digital camera with close up capability (see Note 9).

3. Quartz fluorescence cuvette and holder (see Note 10).

4. ImageJ software (see Note 7).

2.4. Modified TEM assay

1. Copper grids 300-mesh with formvar/carbon coating (e.g., Electron Microscopy Sciences #FCF300-CU).

2. Heavy atom stain (e.g., Uranyl acetate; Electron Microscopy Sciences). Prepare a 2% (w/v) solution in water and filter it through a 0.22 μm pore size syringe filter (e.g., Corning #431229).

3. Filtered ddH₂O (0.22 μm pore size filter).

4. Glow discharge system (e.g., PELCO easiGlow^™ glow discharge cleaning system).

5. Hydrophobic surface (e.g., Parafilm) or silicone grid mat (e.g., Electron Microscopy Sciences #71170).

6. Fine-tipped forceps for grid handling (e.g., Electron Microscopy Sciences #75752–08).

7. Cellulose Blotting paper (e.g., Whatman 3MM Chr; 0.34 mm thick). Cut into small pieces for easiest manipulation.

3. Methods

3.1. Tau Sedimentation Assay

1. Prepare aggregation reactions (composed of tau protein, Aggregation Buffer, inducer and any desired additives) as described previously (29) (see Note 11). Final volumes should be consistent with centrifugation equipment and supplies (see Note 5). Prepare reactions in biological triplicate so that results can be rigorously quantified.

2. At conclusion of incubation period, briefly centrifuge samples to remove condensation (≤1,000g × 10 sec), then transfer to a polycarbonate centrifuge tube and sediment at 100,000×g at 4°C for 1 h.

3. Centrifugation will produce supernatant and pellet fractions (Fig. 2). Carefully remove the supernatant without disturbing the pellet and transfer it to a clean microcentrifuge tube (see Note 12).

4. Resuspend the remaining pellet in Aggregation Buffer (volume equal to starting reaction volume), homogenize by trituration, and then move to a clean tube (see Note 13).

5. Add SDS-PAGE loading buffer to each tube, boil 5 min, then centrifuge briefly to remove condensation. Load equal volumes of all samples onto the prepared SDS-PAGE gels along with the molecular weight ladder (see Note 14).

6. Perform SDS-PAGE and stain gels with Coomassie blue (see Note 15).

7. Image destained gels (see Note 16).

8. Quantify bands by densitometry using the ImageJ gel analyzer tool (see Note 17).

Fig. 2.

Sedimentation assay characteristics. (a) Pellet (white arrows) appearance after centrifugation (100,000×g at 4°C, 1 h) in 7 × 20 mm polycarbonate tubes. Aggregates induced by micellar anionic inducers such as octadecyl sulfate (ODS; 10 μM 2N4R tau, 25:1 inducer:tau molar ratio, 37°C × 16 h incubation) produce white/translucent pellets, whereas those induced by anionic dye molecules such as Geranine G (GG; 10 μM 2N4R tau, 10:1 inducer:tau molar ratio, 37°C × 16 h incubation) stain bright red owing to non-covalent association of dye molecules with aggregates. (b) SDS-PAGE of pellet (P) and supernatant (S) fractions of 2N4R tau incubated in the absence (–) and presence (+) of GG inducer (triplicate assays). The presence of inducer shifts the majority of bulk 2N4R tau from the supernatant into the pellet fraction. (c) Densitometric quantification of 2N4R tau in pellet and supernatant fractions is consistent with the submicromolar minimal concentration determined for 2N4R tau in the presence of GG inducer (15).

3.3. Laser Light Scattering (LLS)

1. Prepare aggregation reactions (composed of tau protein, Aggregation Buffer, inducer and desired additives) as described previously (29). Final volumes should be consistent with cuvette capacity (see Note 18).

2. Set up the apparatus such that the cuvette is in line with the laser and the camera is at an angle of 90° to the incident light (see Note 19).

3. Optimize aperture and exposure time for specific reaction conditions (see Note 20).

4. Capture images at each desired time point/condition (see Note 21).

5. Import images into ImageJ for quantification (see Note 22).

3.4. Electron Microscopy Grid Preparation

1. Prepare aggregation reactions (composed of tau protein, Aggregation Buffer, Inducer and any desired additives) as described previously (29) (see Note 23).

2. Thoroughly clean all tools and surfaces with ethanol. This will help keep the grids from being repelled by the tweezers and keep particulates off the grid.

3. To prepare for glow discharging, carefully place the electron microscopy copper grids onto a clean glass slide, with the carbon coated side facing up (see Note 24).

4. Place the glass slide in the glow discharge and use the auto run function, or other appropriate setting on the glow discharge system (see Note 25).

5. Carefully transfer the glow discharged grids to a silicone mat. Pipette 2–4 μL of sample onto the carbon coated grid surface and allow adsorption for 1 min (see Note 26).

6. Wick away the sample using the corner of a piece of filter paper taking care not to touch the grid.

7. Pipette an equal volume of filtered water onto the grid and remove immediately with filter paper.

8. Pipette an equal volume of 2% uranyl acetate onto the grid and remove immediately with filter paper. Add another equal volume of 2% uranyl acetate and let sit for 1 min. Wick away carefully with filter paper.

9. Wash a final time with an equal volume of water and wick away immediately. Leave on silicone mat and air dry before placing it back in the grid box (see Note 27).

4. Notes

1. Purification protocols are available for full-length, post-translationally modified, and missense and deletion mutant forms of tau (11,30–33).

2. Tau aggregation conditions are inducer dependent. In the presence of Geranine G and other sulfonate dyes, full-length 2N4R tau aggregates efficiently at submicromolar concentrations (10,15), whereas anionic surfactants such as octadecyl sulfate and fatty acids such as arachidonic acid are effective at low micromolar tau concentrations (7,34,35). In contrast, heparin is frequently used above 10 μM tau concentrations (33), whereas spontaneous aggregation of tau deletion mutants corresponding to the microtubule-binding repeat region has been observed at high micromolar concentrations (11,12).

3. These conditions approximate physiological ionic strength, pH and reducing conditions, but the aggregation assays described below are compatible with a wide range of ionic strengths, pHs, and additives, including protease inhibitor cocktails used to protect tau against proteolysis during incubation times extending for days or weeks (e.g., Sigma #P8340).

4. Tau adsorption to surfaces can be minimized by using plasticware that is siliconized or otherwise rated as low protein binding. This includes pipette tips used for transferring tau-containing solutions and tubes or 96-well plates used for incubations. Tubes and plates should be sealed to avoid evaporation during incubation.

5. Thick-wall polycarbonate tubes compatible with 100 – 200 μL volumes and ≥100,000×g forces work well in this assay (e.g., Beckman #343775; 230 μL nominal capacity). Throughput can be maximized by pairing these tubes with rotors having high tube cavity numbers (e.g., Beckman Type 42.2 Ti; 72 tube cavities).

6. Standard Laemmli slab gel and Coomassie blue staining protocols are compatible with this assay (36,37).

7. ImageJ is freeware developed by Wayne Rasband at the National Institutes of Health (38,39). It is available for download at: https://imagej.nih.gov/ij/

8. Laser light sources ranging from 457 – 635 nm in wavelength and 5 – 100 mW in power output have been used successfully in this assay.

9. Monochrome CMOS cameras (e.g., Imagingsource #DMK 22BUC03) conveniently provide USB connectivity to laboratory computers. Fitting the camera with a close focusing macro video lens (e.g., 12.5 – 75 mm F 1.8 Macro) equipped with close up filters (≥ +6 total diopter strength) allows capture of light scattering images at distances as close as 1 cm.

10. Fluorescence cuvettes have four optical sides and are available in a broad range of sizes. Micro (350 – 700 μL capacity) and submicro (10 – 350 μL capacity) quartz cuvettes are appropriate for this assay.

11. Sedimentation is broadly compatible with most inducers (Table 1). Anionic microspheres are an exception because they pellet along with aggregates during centrifugation. As a result, tau quantified in the pellet fraction will contain both filamentous aggregates as well as all forms of tau bound to bead surfaces.

12. Although tau aggregate pellets are easily visualized when prepared from above 10 μM tau concentrations or in the presence of colored inducers such as Geranine G (Fig. 2), they can be less distinct at low tau concentrations. To anticipate the location of these pellets, mark the side of the ultracentrifuge tube pointing away from the center of the rotor prior to centrifugation.

13. Triturate forcefully for 1 min or until pellet is fully homogenized.

14. Quantification of Coomassie blue-stained bands on SDS-PAGE tend to be linear though microgram quantities of protein (40). Nonetheless, linearity should be established locally, and the amounts of tau loaded on each well should be within the validated linear range while not exceeding the maximum volume of the gel wells. Typically, total tau amounts of 1 μg are appropriate. For example, 5 μL aliquots of fractions containing 5 μM tau will likely be within the linear range of quantification.

15. All full-length tau isoforms run anomalously on SDS-PAGE (41). For example, the largest tau isoform (2N4R) migrates relative to standards near 66 kDa, which is well above its calculated value.

16. Standard office scanners and other image detectors can generate images of sufficient quality for quantitative densitometry (42).

17. To quantify sedimentation results, band densities should be normalized to a standard within each gel (e.g., a known amount of unaggregated tau monomer), or by directly comparing supernatant and pellet fractions on the same gel.

18. Assay additives can complicate LLS assays. First, micellizing species such as detergents can contribute to light scattering signal. This can be detected and controlled by using aggregation-incompetent tau (e.g., 2N4R^I277P/I308P; (43,44)) in parallel reactions and subtracting the resulting signal (27,45). Second, small molecules that absorb visible light at laser wavelengths can directly interfere with LLS signal (e.g., dye molecules). This problem can be avoided in some instances by using a laser wavelength outside the absorbance optimum of the added compound. Finally, bubbles and large particulates such as dust interfere with LLS. Therefore, reagents should be filtered (0.22 μm filter) and degassed before use.

19. This orientation eliminates the need to block extraneous scattering from angles other than 90° (26).

20. LLS intensity is estimated from images taken by the digital camera in the form of the average gray scale value in a pixel region of interest. The camera aperture and shutter speed must be set so that this value is not saturated and lies on the linear region of the gray scale. This calibration is conveniently done using images from pilot aggregation reactions at plateau, which will identify the highest scattering intensities encountered in data collection as well as quantify a blank or “dark current” that should be subtracted from all measurements (27).

21. Dynamic range can be maximized by normalizing for exposure time (27).

22. Synthetic tau filaments display minimal angular dependence of light scattering intensity (26), consistent with their flexible nature (e.g., as observed by electron microscopy (7,46)). Consistent with this observation, the intensity of LLS by synthetic tau filaments closely tracks length measurements made by TEM through all phases of the aggregation time course (27). As a result, LLS can be used to quantify aggregation dynamics, including comparisons of lag times and minimal concentrations (27,47).

23. Because of its ability to distinguish filamentous morphology, TEM is compatible with most additives. For example, it is possible to accurately measure tau filament lengths in the presence of lipid vesicle and anionic microsphere inducers that appear alongside filaments in micrographs (7,48).

24. The carbon coated side of the grid has a “shiny” or “smooth” appearance. It also appears darker and more reflective than the non-coated side. Some grids have additional measures to identify the coated side when viewed by TEM.

25. Glow discharging hydrophilizes and cleans the grids, promoting the even spreading of fibrils across their surfaces. Most importantly, it also facilitates adsorption of filaments composed of tau fragments (e.g., K18 tau) to the grid. However, it is not necessary when using full-length tau fibrils.

26. The adsorption time for sample on the grid can be increased if there are not enough fibrils when imaged on EM. However, adsorption time must be the same for any grids that are quantitatively compared.

27. The method presented herein decreases the amount of sample used and grid manipulation. The approach described previously of inverting the grid onto a droplet of sample (9) is a viable alternative.