Amyloid β-protein oligomers promote the uptake of tau fibril seeds potentiating intracellular tau aggregation

To determine which forms of self-assembled or aggregated Aβ might promote tau seeding, we first expressed and purified recombinant Aβ and tau and prepared different self-assembled forms of each. Recombinant Aβ (1–42) (Fig. 1a) was expressed and purified as described previously [32]. To eliminate pre-formed aggregates during purification, the protein was fractionated using size-exclusion chromatography (SEC), yielding a major peak with an apparent retention time of a trimer (Additional file 1: Figure S1a), which previously has been shown to consist of a mixture of monomers and small oligomers, defined as low molecular weight (LMW) Aβ [33]. This LMW Aβ showed a major band when analyzed by native-PAGE at ~ 20 kDa and a smeary band consistent with molecular weights ranging from 70 to 200 kDa (Additional file 1: Figure S1b), likely representing the immediate self-assembly of Aβ in the PBS solution. Examining the preparation by electron microscopy (EM) showed occasional spherical structures, but mostly the species in freshly prepared Aβ were too small to be observed by EM (Additional file 1: Figure S1c). The freshly prepared Aβ was then quiescently incubated at 37 °C at 10 μM in PBS to generate different self-assembly states. We monitored the self-assembly of Aβ by thioflavin T (ThT) fluorescence (Fig. 1c). After incubation for 18 h, native PAGE showed a smeary band consistent with molecular weight ≥ 250 kDa, suggesting the presence of high molecular weight oligomers composed of ≥ 55 copies of the Aβ peptide (~ 4.5 kDa per monomer) (Additional file 1: Figure S1b). EM images of these samples, termed oligomeric Aβ, showed abundant quasi-spherical structures with diameters ranging between 20 and 60 nm (Additional file 1: Figure S1d,e). Finally, after ~ 100 h of incubation, Aβ formed typical, unbranched amyloid fibrils (Fig. 1c). The native PAGE lane for the fibril sample shows no bands (Additional file 1: Figure S1b), indicating a complete conversion of the samples into insoluble fibrils unable to migrate on the gel, which is also confirmed by the plateau in the ThT kinetics (Fig. 1c).

Recombinant tau repeat domain (RD, Fig. 1b) was used to generate fibrillar seeds in PBS containing 20 mM DTT at 37 °C. We did not use aggregation inducers, such as heparin, because of their potential effect on intracellular tau seeding/aggregation, complicating data interpretation. Under these conditions, following a lag phase of ~ 30 h, a sharp increase in ThT fluorescence was observed, suggesting tau fibril formation. In agreement with these results, EM images showed no aggregates at 0-h incubation and elongated fibrils after 100 h of incubation (Fig. 1d). In contrast to Aβ, for which quasi-spherical oligomers were observed at intermediate time points (from 10 to 30 h), we did not observe oligomers of tau RD by EM.

Following the preparation of different aggregated species of Aβ, we assessed their impact on tau seeding using the HEK293T tau biosensor cell line developed by Diamond and co-workers for monitoring and quantifying intracellular tau aggregation and seeding [34, 35]. Expressing tau RD containing the disease-associated P301S substitution fused to either CFP or YFP, the biosensor cells produce a FRET signal upon aggregation of tau, which can be quantified using flow cytometry. As a control, we first tested the effect of Aβ on the biosensor cells in the absence of tau seeds. The cells were treated with up to 200 nM freshly prepared, oligomeric, or fibrillar Aβ without the addition of transfection reagent, and FRET was quantified by flow cytometry. As expected, minimal tau aggregation (integrated FRET density (IFD): < 10, Additional file 1: Figure S2) was observed after 48 h, and even after 2 weeks (Additional file 1: Figure S3), demonstrating that neither form of Aβ induced tau aggregation without the addition of tau seeds in these cells. Our additional control experiments using tau RD seeds on their own, again not mixed with transfection reagents or any other additive, showed no measurable intracellular aggregation in the biosensor cells (Fig. 2a, b).

Next, to test whether Aβ facilitates tau seeding, biosensor cells were seeded with sonicated tau RD fibrils 24 h after adding freshly prepared, oligomeric, or fibrillar Aβ, and the resulting levels of intracellular tau aggregation were quantified after an additional 48-h incubation. Treating cells with 100–500 nM Aβ oligomers 24 h before addition of tau RD seeds led to a significant, dose-dependent increase in intracellular tau aggregation. The aggregates were visualized as bright puncta by fluorescence microscopy (Fig. 2a, white arrows) and were quantified by flow cytometry (Fig. 2b). Treatment of the cells with Aβ fibrils also promoted tau aggregation, yet to a substantially lower extent (~ 10%) than Aβ oligomers. Treatment with freshly prepared Aβ or a mixture of freshly prepared Aβ and Aβ fibrils did not promote tau aggregation.

We further asked whether Aβ oligomers could facilitate intracellular tau aggregation seeded not only by the tau RD, but also by full-length recombinant human tau (tau40) or by brain extracts from transgenic PS19 mice expressing human 1N4R tau containing the P301S substitution [36]. Addition of tau40 fibrils at 100–500 nM induced a low level of tau aggregation in the biosensor cells (Fig. 2c, blue curve). Pre-treatment of the cells with 500 nM Aβ oligomers doubled the seeding effect of tau40 fibrils (red curve). Following the same trend, pre-treatment with Aβ oligomer led to a small increase in seeding by freshly prepared tau40 (green and black curves). When we quantified tau aggregation seeded by PS19-mouse brain extracts, containing human 1N4R P301S tau, priming with 200 or 500 nM Aβ oligomers dose-dependently increased tau seeding by 15 and > 30 fold, respectively, compared to brain extract added without Aβ pre-treatment (Fig. 2d). Representative fluorescence microscopic images of these cells are shown in Additional file 1: Figure S4. Though tau RD, tau40, and PS19 brain extracts had different seeding capacities, in all cases, priming with Aβ oligomers substantially enhanced the seeding of tau.

We also assessed whether the priming of the biosensor cells by Aβ oligomers required a preincubation time, or could be achieved immediately. When tau seeds were added to the cells simultaneously with Aβ oligomers, we did not observe promotion of tau seeding (Additional file 1: Figure S5), suggesting that sufficient incubation time (~ 24 h) of the pretreatment was needed for Aβ oligomer to prime the cells for enhancing tau seeding.

An important mechanistic question is whether Aβ oligomers promote seeding of tau specifically or could facilitate seeding of other amyloidogenic proteins in a non-specific manner. To address this question, we examined the effect of Aβ oligomers on seeding of α-synuclein, whose aggregation has been associated with Parkinson’s disease and other synucleinopathies [37]. Similar to tau, α-synuclein pathology has been reported to accumulate through brain regions, and α-synuclein aggregates often are found in the AD brain [38]. Using HEK293T biosensor cells co-expressing CFP- or YFP-fused A53T-α-synuclein, we added up to 1 μM sonicated α-synuclein fibrils in the absence or presence of Aβ-oligomer priming. In contrast to the experiments using tau seeds and tau-biosensor cells, we did not observe an enhancement in the seeding of α-synuclein in α-synuclein-biosensor cells pre-treated with up to 200 nM Aβ oligomers compared to cells that were not pre-treated (Additional file 1: Figure S6), suggesting that the seeding enhancement by Aβ oligomers is specific to tau.

The FRET-based biosensor cells are a convenient tool for sensitive quantification of tau or α-synuclein seeding, but may not fully represent seeding in neurons. Therefore, we sought to determine whether Aβ oligomers facilitate tau seeding in primary mouse neurons and a human neuroblastoma cell line. Primary hippocampal neurons were isolated from the PS19 mice and treated with tau seeds with or without Aβ-oligomer priming. Because this cellular system does not express fluorescently labeled tau and is not applicable to FRET-based flow cytometry, we used confocal microscopy instead to qualitatively assess seeding (Fig. 3a). When immunostained with antibodies HT7 or AT8, which recognize total human tau and phosphorylated tau (at Ser 202 and Thr 205), respectively [39], primary hippocampal neurons treated with 500 nM tau RD seeds alone showed a few bright fluorescent puncta, while untreated cells showed no such puncta. Treatment with 200 nM Aβ oligomers 24 h before the addition of tau seeds led to a marked increase in red fluorescent puncta (indicated by white arrows, Fig. 3a, bottom panels).

To quantitatively measure tau seeding in neuronal cells, we established a cellular assay in which human SH-SY5Y neuroblastoma cells were incubated with tau seeds with or without pre-treatment with Aβ oligomers. The cells were washed extensively with PBS to remove any tau and Aβ that might have been present in the culture media, lysed, and their lysates centrifuged to remove insoluble material and cell debris. The clarified cell lysates (tau concentration of 26 nM, measured by Western blot and ELISA, Additional file 1: Figures S7-S8) were added to the tau biosensor cells for measurement of seeding (Fig. 3b). Priming the SH-SY5Y cells with Aβ oligomers yielded > 4-fold increase in tau aggregation in biosensor cells compared to the SH-SY5Y cells without Aβ-oligomer priming (Fig. 3c).

To visualize the cellular uptake of tau seeds, we conjugated fluorescein to tau RD (FITC-tau), allowed the protein to form fibrils, and prepared seeds. We then primed SH-SY5Y cells with 200 nM freshly prepared, oligomeric, or fibrillar Aβ for 24 h, followed by seeding with 500 nM FITC-tau fibrils. After extensive washing with PBS, the cells were inspected by fluorescence microscopy (Fig. 4a) and seeding was quantified using flow cytometry (Fig. 4b). Without Aβ-oligomer priming, 12 ± 3% of the SH-SY5Y cells showed internalized FITC-tau seeds. Treatment with freshly prepared Aβ (12 ± 4%) or Aβ fibrils (16 ± 3%) did not affect FITC-tau internalization. Priming with oligomeric Aβ increased the uptake of FITC-tau seeds significantly to 64 ± 8%, suggesting that Aβ oligomers facilitated tau seeding by promoting the internalization of the seeds. Aβ-promoted internalization of tau seeds was also observed in wild-type mouse primary neurons (Additional file 1: Figure S9).

Lastly, to investigate whether inhibition of tau seed uptake could block Aβ promoted tau seeding, we tested heparin, a known tau uptake inhibitor [40], in the tau biosensor cells. Flow-cytometry analysis revealed that increasing concentrations of heparin, 1 to 50 μM, decreased the Aβ-promoted tau seeding in a dose-dependent manner (Fig. 4c). We confirmed that heparin inhibits tau-seed uptake using wild-type HEK293T cells instead of the biosensor cells because of spectral overlap between the fluorescence of FITC and YFP/CFP. Similar to the SH-SY5Y cells, Aβ oligomers increased the internalization of FITC-tau seeds in wild-type HEK293T cells > 5-fold compared to untreated cells (Fig. 4d). Addition of 10 μM heparin reduced tau internalization to the level of non-primed cells.

Animal care was conducted in compliance with the US Public Health Service Guide for the Care and Use of Laboratory Animals, and the procedures were approved by the Institutional Animal Care and Use Committee at the University of California, Los Angeles. Eight-month-old P301S (PS19) transgenic mice on (C57BL/6 x C3H) F1 background and wild-type mice were used for the study.

Aβ (1–42) was expressed and purified as described previously [32]. Briefly, Escherichia coli BL21 plyS (DE3) cells were transformed with a plasmid encoding Aβ (1–42) conjugated to maltose binding protein (pET28a-MBP-Aβ42). The expression of the fusion protein MBP-Aβ42 was induced with 1 mM of isopropyl-β-D-thiogalactopyranoside (IPTG) at 30 °C for 4 h. The culture was harvested and washed. The denatured fusion protein was purified using a HisTrap HP column (GE healthcare). After overnight dialysis, the fusion protein was cleaved overnight by TEV protease. The uncleaved protein and TEV were removed using Ni²⁺ affinity chromatography. The Aβ (1–42) in the flow-through was purified by RP-HPLC and lyophilized. To ensure the homogeneity of the Aβ sample, the lyophilized powder was dissolved in 60 mM NaOH and fractionated by SEC (GE, Superdex 200 Increase) in 20 mM NaOH to eliminate pre-formed Aβ aggregates. The final concentration was determined by a BCA assay, and the stock was stored in − 80 °C.

Human tau RD (residues 244–372) and full-length tau (tau40, residues 1–441) were expressed and purified as described previously [56]. Briefly, for the expression of tau RD, BL21 (DE3 GOLD) competent cells and a pNG2 vector was used. Protein expression was induced by addition of 0.5 mM IPTG for 3 h at 37 °C, and the cells were lysed by sonication. Cell lysates were then boiled for 20 min and centrifuged to remove all insoluble proteins. The remaining soluble protein was purified using a HighTrap SP ion-exchange column (GE Healthcare), and fractions were analyzed by SDS-PAGE/Coomassie blue staining. Enriched fractions were further purified using a HiLoad 16/600 Superdex 75 SEC column (GE Healthcare). Purified tau was concentrated to 50–100 mg/ml and stored at − 80 °C.

For the expression of human tau40, a C-terminal His-tag was used. After IPTG induction, the cells were harvested and collected. The protein was purified using a HisTrap HP column and then purified by ion-exchange chromatography and SEC as described for tau RD.

Before the ThT kinetics assay, freshly SEC-fractionated Aβ was mixed with 60 mM HCl to neutralize the NaOH in the stock solution. After 3-min sonication, the Aβ sample was diluted in PBS to a final concentration of 10 μM and 30 μM ThT (CalbioChem) was added. The reaction mixture was filtered through a 0.2-μm filter, split into 3–4 replicates, and immediately placed in a Corning 96-well Nonbinding plate (black, nonbinding surface microplate). The ThT fluorescence signal was measured every 5 min in quiescent conditions using a Fluostar Omega plate reader (BMG Labtech, Offenburg, Germany) with excitation and emission wavelengths of 440 and 490 nm, respectively, at 37 °C.

Freshly purified tau RD and tau40 were diluted in PBS containing 2 mM DTT, 40 μM ThT to a final concentration of 250 μM in the absence of any aggregation inducers (i.e., heparin). The reaction mixture was split into 3–4 replicates and placed in 96-well plates (Corning 3881). ThT fluorescence intensity was measured every 15 min with double orbital shaking at 37 °C in the same plate reader until a plateau was reached.

Tau RD P301S FRET Biosensor (ATCC CRL-3275) cells were cultured and analyzed as described previously [31]. The cells were grown in DMEM (Dulbecco’s modifications of Eagle’s medium with l-glutamine and 4.5 g/l glucose) supplemented with fetal bovine serum (FBS), 100 units/ml of penicillin G, and 0.1 mg/ml of streptomycin sulfate, in a humidified atmosphere of a 5% CO₂ at 37 °C. Trypsin-treated HEK293T cells were plated on collagen-coated flat 96-well plates at a density of 2.5 × 10⁴ cells/well in 200-μl culture medium and incubated at 37 °C in 5% CO₂. After 24 h, the cells were treated with different self-assembly states of Aβ or with media alone. Tau fibrils (tau RD, tau 40 and mouse brain extract) were diluted with Opti-MEM (GIBCO) and sonicated for 10 min in an ultrasonic water bath. Twenty-four hours after the treatment with Aβ, the culture media were replaced with fresh media containing the tau seeds. Tau aggregation in the biosensor cells was visualized at 72 h by fluorescence microscopy using green channel (ex: 485; em: 520). At 96 h, the cells were harvested after extensive washing and treatment with trypsin. The harvested cells were prepared in 200 μl chilled buffer (HBSS, 1% FBS, 1 mM EDTA) and then stored at 4 °C until they were analyzed by FRET-based flow cytometry.

α-synuclein biosensor cells (a gift from Marc Diamond, UTSW) [30] were used for the study. Cells were grown in same DMEM growth media and 2.5 × 10⁴ cells were plated on collagen-coated flat 96-well plates. After 24 h, the cells were treated with 200 nM Aβ oligomers followed by addition of sonicated α-synuclein fibrils at 48 h. The preparation of recombinant α-synuclein fibrils was performed as described previously [57]. α-Synuclein seeding was visualized at 72 h by florescence microscopy, and at 96 h, the cells were harvested and prepared for FRET-based flow cytometry.

Intracellular protein aggregation of tau or α-synuclein was quantified by FRET-based flow cytometry. The protocol was adapted from reference [31]. All experiments were performed using a Digital Analyzers LSRII (IMED) flow cytometer. The fluorescence intensities of the FRET pair (ex: 405 nm; em: 525/50 nm), CFP-fusion proteins (ex: 405 nm; em: 450/50 nm), and YFP-fusion proteins (ex: 488 nm; em: 525/50 nm) alone were measured. For each experiment, FRET signals of 20,000 cells per replicate were analyzed to differentiate the aggregated protein from the non-aggregated protein. FRET gating was introduced to exclude all of the FRET-negative cells and include the FRET-positive cells. Integrated FRET density (IFD), defined as the percentage of FRET-positive cells multiplied by the median fluorescence intensity of FRET-positive cells, was calculated for all analyses. All experimental data were analyzed using GraphPad Prism 7.0. Plots were fitted to a non-linear sigmoidal curve. The flow-cytometry quantification of protein aggregation was conducted for a minimum of three independent experiments with at least three replicates in each experimental condition.

Human neuroblastoma SH-SY5Y cells were cultured in Iscove’s modified Dulbecco’s medium supplemented with 15% fetal bovine serum (FBS), 1% l-glutamine, and 1% penicillin/streptomycin. Cells were transferred to collagen-coated flat 6-well plates at a density of 3.0 × 10⁶ cells/well in 1-ml culture medium and incubated at 37 °C in 5% CO₂. The cells were treated with Aβ oligomers at 24 h, followed by tau seeds at 48 h. At 72 h, the cells were lysed by sonication in 15% RIPA buffer and the lysate supernatants were harvested. ELISA was used for quantifying total tau concentration in the lysate supernatants. After normalization of tau concentration, the lysate supernatant was added to tau biosensor cells to measure its seeding capacity. After 48 h, the HEK293T cells were collected, and intracellular tau aggregation was quantified using FRET-based flow cytometry.

Freshly purified tau protein (50 μM in PBS containing 20 mM DTT) was incubated with shaking at 37 °C for > 100 h until most of the protein was converted into fibrils (a plateau of ThT fluorescent signal monitoring tau aggregation was reached). The tau fibrils were then diluted in a reaction buffer containing 5 μM DTT, 10 mM HEPES, pH 7.4, and 100 mM NaCl to a final concentration of 8 μM. 0.025 mg of Alexa Fluor 488 NHS Ester per 200 μL was added into the reaction mixture and incubated for 1 h at room temperature and then overnight at 4 °C with end-over-end rotation. To stop the reaction, the unconjugated dye was quenched with 100 mM glycine in PBS for 1 h. The labeled fibrils were washed extensively with PBS buffer by filtration using a molecular weight cut-off of 10 kDa.

SH-SY5Y or HEK293T cells were used to measure cellular uptake of tau fibril seeds. 3 × 10⁴ cells were plated in collagen-coated, flat 96-well plates, incubated for 24 h, and then treated with freshly prepared Aβ, Aβ oligomers, or Aβ fibrils. Fluorescently labeled FITC-tau was diluted in Opti-MEM (GIBCO) and added to the cells after an additional 24-h incubation. Tau uptake was monitored using florescence microscopy (ex: 485 em: 520) 24 h later. After an additional 24-h incubation, cells were harvested after extensive washing with 1× PBS and flow cytometry was used to quantify the cellular uptake of tau fibril seeds by calculating the percentage of FITC-positive cells.

Five-microliter aliquots from aggregation reactions were taken at different time points and applied to carbon-coated 400-mesh Formvar grids (Electron Microscopy Science), which had been glow-discharged using a Pelco Easy-Glow unit for 2 min immediately before applying the samples. The samples were wicked off after 1 min, stained with 2% uranyl acetate for 2 min, and analyzed using a JEOL JEM1200-EX transmission electron microscope.

Cells were lysed by sonication in mild (15%) RIPA buffer, and the lysate supernatants were harvested by centrifugation for 20 min at 14,000 rpm in 4 °C. The collected supernatants were dissolved in LDS sample buffer (Thermo Fisher Scientific) and boiled for 5 min prior to loading on NuPAGE 4–12% Bis-Tris polyacrylamide gels. Following fractionation by SDS–PAGE, proteins were transferred to polyvinylidene difluoride membranes and blocked with 5% non-fat dry milk in Tris-buffered saline containing 0.1% Tween 20 (TBST) for 60 min at room temperature. The membranes were incubated for 2 h at room temperature with anti-human tau monoclonal antibody HT7 (Cell Signaling Technology, Danvers, MA) at a 1:1000 dilution. After 3-min washes with TBST, membranes were incubated with HRP-conjugated [donkey] anti-mouse secondary Anti-thymocyte globulin (Zymed, San Francisco, CA) at a 1:5000 dilution and washed three times again with TBST. The transferred proteins were visualized using an enhanced chemiluminescence (ECL) detection kit (HyGLO Quick Spray Kit, Denville Scientific).

Tau concentration was quantified using an anti-tau ELISA kit (Invitrogen, cat# KHB0041) according to the manufacturer’s instructions. After cell lysis as described above, the resulting supernatant was collected for analysis. Fifty microliters of each sample was analyzed in triplicates.

Twelve-millimeter coverslips were placed into the wells of 24-well plates and coated with 500 μl 0.5 mg/ml poly-ornithine dissolved in 50 mM sodium tetraborate, pH 8.3, overnight at 4 °C. After rinsing with PBS, pH 7.4, the coverslips were coated with 5 μg/ml laminin dissolved in PBS for 2–3 h at 37 °C and then rinsed and stored in PBS at 4 °C. Pyramidal hippocampal neurons were isolated and cultured from P0 or P1 postnatal P301S mice (PS19 transgenic mouse line, Jackson Laboratories) as described previously [58]. Briefly, the hippocampi were dissected under a microscope and collected into a 15-ml tube using ice-cold PGB buffer containing 0.2 g BSA, 0.9 g glucose in 200 ml PBS, pH 7.4. The PGB buffer was replaced with a solution of 10 ml, 0.5 mg/ml papain, and 0.6 μg/ml DNAase (Sigma) in PGB, and the tissue was incubated for 20 min at 37 °C, rinsed twice with PGB buffer, and dissociated in 2 ml PGB buffer by trituration using a flame-polished, plugged glass pipette. Then, 8 ml of PGB buffer was added and the preparation was centrifuged at 400×g for 10 min. The supernatants were discarded, and 2 mL of complete medium (0.5 mM glutamine, 100× penicillin/streptomycin, 50× B-27 in neurobasal medium, Gibco) was added. The cell pellet was dissociated gently using a flame-polished glass pipette, and the cells were counted using a hemocytometer. Seventy thousand to 100,000 cells per well were plated in 0.5 mL of the complete medium on the previously prepared coverslips. The cells were maintained at 37 °C in a 5% CO₂ atmosphere. Half of the medium was changed 1–2 times a week.

On day 8 of the primary neuron culture, 200 nM Aβ oligomers in complete medium were added. Twenty-four hours later, 500 nM sonicated tau RD fibrils were added. After an additional 24-h incubation, the cells were fixed in 4% (v/v) paraformaldehyde in PBS for 15 min, washed three times in PBS, and permeated for 1 h at room temperature in blocking buffer containing 0.1% BSA, 5% donkey serum, 0.3% Tween-20 in PBS, pH 7.4. The blocking buffer then was removed, and the cells were stained with HT7 or anti-phosphorylated tau monoclonal antibody AT8 (ThermoFisher) at 1:1000 dilution in blocking buffer overnight at 4 °C, followed by Alexa-Fluor-555-conjugated donkey anti-mouse secondary antibody (ThermoFisher Scientific) at 1:500 dilution in blocking buffer for 1 h at room temperature. After washing in PBS, the coverslips were mounted on microscope slides using Prolong Gold Antifade reagent with DAPI (Thermo Fisher Scientific) and imaged using a Leica SP8-SMD Confocal Laser Scanning Microscopy Platform.

Statistical analyses for two-group comparison were performed by t test calculator (QuickCalcs, GraphPad) with the choice of unpaired t test. For multiple comparison, ordinary one-way ANOVA multiple comparisons were performed using GraphPad Prism software ver. 7.0 (San Diego, CA). All graphs were generated using Prism software.