Recombinant Tau Purification and in Vitro Fibrillization

The cDNAs coding for 1) the longest isoform of Tau (2N4R) with a Myc tag at the 3′ end (3′Myc-T40), 2) truncated Tau containing only four MT-binding repeats with a Myc tag at the 5′ end (5′Myc-K18), and 3) 5′Myc-K18 containing P301L mutation (5′Myc-K18/P301L) were cloned into NdeI/EcoRI sites in pRK172 bacterial expression vector. Each protein was expressed in BL21 (DE3) RIL cells and purified by cationic exchange using a fast protein liquid chromatography (FPLC) column as described previously (34). In vitro fibrillization was conducted by mixing 40 μm recombinant Tau with 40 μm low molecular weight heparin and 2 mm DTT in 100 mm sodium acetate buffer (pH 7.0). Except 3′Myc-T40, which was agitated at 1,000 rpm for 5 days, the other proteins were incubated at 37 °C without shaking for 2–3 days. Successful fibrillization was confirmed using thioflavin T fluorescence assay, sedimentation tests, and negative stain electron microscopy (supplemental Fig. S1) as described previously (34). Before being used for transduction on cells, fibrillization mixture was centrifuged at 100,000 × g for 30 min at 4 °C. The resulting pellet was resuspended in equal volume of 100 mm sodium acetate buffer (pH 7.0), without heparin and DTT, and frozen as single use aliquots at −80 °C.

Cell Culture and Fibril Transduction

QBI-293 cells (QBiogene) were grown in full media (DMEM, 10% FBS) supplemented with penicillin/streptomycin and l-glutamine. For fibril transduction, cells were plated at 400,000/well in a 6-well tissue culture plate 1 day before transient transfection with WT or mutant Tau in pcDNA5/TO or pcDNA3.1(+) using FuGENE® 6 (Roche Applied Science) as per manufacturer's instructions at 2 μg of DNA per well. About 16 h after transfection, resuspended fibril aliquots were diluted into 10 μm using 100 mm sodium acetate buffer (pH 7.0) and sonicated with 20–30 pulses, 80 μl of which was added to one tube of BioPORTER reagent, gently vortexed for 5 s, and allowed to stand at room temperature for 5–10 min. During the formation of the pffs-reagent complex, cells were washed once with 2 ml of OptiMEM (Invitrogen) and placed on 500 μl of OptiMEM before the fibril-reagent complex was diluted with 420 μl of OptiMEM and added to cells. For experiments performed on cells plated on 12-well plates, all reagents were reduced by half. Six to seven hours after the addition of fibrils, cells were placed on starvation medium (DMEM, 0.5% FBS, penicillin/streptomycin, l-glutamine) to reduce cell division, 24 h after which they were washed once with Versene/EDTA, incubated with 0.05% trypsin/EDTA for 10 min at 37 °C to remove the majority of membrane-associated fibrils, and replated to poly-d-lysine-coated glass coverslips for immunostaining. Fixing was performed one night after replating.

For fibril alone transduction without BioPORTER reagent, cells were placed on starvation medium right before the addition of fibrils. Fibrils were processed in exactly the same way, and the same amount was used as in reagent-mediated transduction. Cells were trypsinized and replated either 4 or 30 h after the initial addition of fibrils. For experiments using WGA, different doses of WGA were added to starvation medium together with pffs. To inhibit endocytosis of WGA, 0.1 m GlcNAc was incubated with medium containing pffs and 10 μg/ml WGA for 1 h at 37 °C before being added to cells that were preincubated in 0.1 m GlcNAc-containing medium for 1 h. Cells were placed on WGA-free starvation medium 5–6 h after fibril addition and fixed after overnight incubation to minimize possible toxicity resulted from WGA treatment. No replating was done for biochemical analysis except for experiments attempting to quantify cell-associated pffs (Fig. 7F) and 4 h fibril-alone transduction experiments (Fig. 7C) in which cells were transferred to new 6-well plates after the incubation period.

Immunocytochemistry

Immunocytochemistry was performed 48 h after the initial addition of fibrils unless indicated otherwise. Cells were fixed in 4% paraformaldehyde for 10 min and permeabilized with 0.1% Triton X-100 for 15 min or fixed with 4% paraformaldehyde containing 1% Triton X-100 for 15 min to remove soluble proteins. For visualization of microtubules, cells were fixed with 0.3% glutaraldehyde in PEM buffer (80 mm PIPES (pH 6.8), 5 mm EGTA, 1 mm MgCl₂) for 10 min, extracted with 1% Triton X-100 for 15 min before being quenched with 10 mg/ml sodium borohydride for 7 min, followed by 0.1 m glycine for 20 min. After blocking with 3% BSA and 3% FBS for at least 1 h at room temperature, cells were incubated with specific primary antibodies (see supplemental Table 1) for 2 h at room temperature or overnight at 4 °C followed by staining with appropriate Alexa Fluor 594- or 488-conjugated secondary antibodies (goat anti-mouse/rabbit/mouse IgG1/mouse IgG2b) for 1–3 h at room temperature. DAPI was added to PBS wash to label cell nuclei. For ThS staining, cells were incubated with 0.005% ThS for 8 min and differentiated with 70% ethanol four times for 5 min each.

Two-stage immunostaining was employed to distinguish between membrane-associated and truly internalized pffs. Live cells were incubated with rabbit polyclonal anti-Myc Ab diluted in medium containing 25 mm HEPES for 1 h at 4 °C. After three washes with cold medium, cells were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100. Following blocking, cells were incubated with monoclonal anti-Myc Ab 9E10 for 1–2 h at room temperature. Rabbit Myc Ab is expected to bind to cell membrane-associated pffs only, where 9E10 would bind to both membrane-associated and intracellular pffs; therefore, truly internalized pffs should only be recognized by 9E10 but not rabbit Myc Ab. Epifluorescence images were acquired using Olympus BX 51 microscope (B&B Microscopes) equipped with a digital camera DP71 (Olympus) and DP manager (Olympus).

Sequential Extraction and Western Blot

Cells were washed once with PBS before being scraped into Triton lysis buffer (1% Triton X-100 in 50 mm Tris, 150 mm NaCl (pH 7.6)) or Sarkosyl lysis buffer (1% Sarkosyl in 50 mm Tris, 150 mm NaCl (pH 7.6)) containing protease and phosphatase inhibitors and incubated on ice for 10–15 min. For Sarkosyl extraction, lysates were rotated at room temperature for 1 h before the next step. Following sonication, lysates were centrifuged at 100,000 × g for 30 min at 4 °C. Supernatants were kept as “Triton/Sarkosyl fraction,” whereas pellets were washed once in Triton/Sarkosyl lysis buffer, resuspended, and sonicated in SDS lysis buffer (1% SDS in 50 mm Tris, 150 mm NaCl (pH 7.6)). After centrifugation at 100,000 × g for 30 min at 22 °C, supernatants were saved as “SDS fraction.” Equal proportions of Triton/Sarkosyl and SDS fractions were resolved on 10% SDS-polyacrylamide gels, transferred to nitrocellulose membranes, and blocked in 5% milk in TBS before probing with specific Abs (see supplemental Table 1). Protein-free block (ThermoFisher) was used for PHF-1 blots.

Pulse-Chase

Reagent-mediated transduction was performed as described above. One day after transduction, cells were placed on methionine-free medium for 15 min and then pulsed with methionine-free medium containing [³⁵S]methionine (150 μCi for each well on 12-well plate) for 20 min, following which cells were switched into complete medium containing 5 mm cold methionine and chased for different durations. During harvest, cells were washed once in Hanks' balanced salt solution and sequentially extracted using 1% Triton X-100 followed by 1% SDS lysis buffer. Tau was immunoprecipitated from cell lysates using a combination of mAbs T46 and Tau 5.

Electron Microscopy (EM)

To visualize pffs made in vitro, 6 μl of unsonicated or sonicated fibrils was placed on 300 meshed Formvar/carbon film-coated copper grids (Electron Microscopy Sciences) for 3 min, washed twice with 50 mm Tris buffer (pH 7.6) for 5 min each, stained with 10–15 drops of 2% uranyl acetate, and visualized with Jeol 1010 transmission electron microscope (Peabody). Immuno-EM of reagent alone-treated cells and fibril-transduced cells were performed as described previously (30) using PHF-1 as the primary antibody.

Sandwich ELISA for Quantifying Acetylated Tubulin and Tau in Cell Lysates

Levels of acetylated tubulin (Ac-tub) and Tau in cell lysates were quantified using sandwich ELISA in 384-well plate format. For Ac-tub ELISA, tubulin mAb 12G10 was used as the capture antibody, and horseradish peroxidase (HRP)-conjugated Ac-tub mAb was used as the reporting antibody. Standard curves were generated using serially diluted brain lysates from WT mice. 18 μg of cell lysates diluted in EC buffer (0.02 m sodium phosphate buffer (pH 7.0), 2 mm EDTA, 0.4 m NaCl, 0.2% BSA, 0.05% CHAPS, 0.4% Blockace, 0.05% NaN₃) was loaded to each well and incubated overnight at 4 °C. The following day, after 4 h of incubation with reporting antibody at room temperature, trimethylbenzidine peroxidase substrate solution was added, and the plate was read using SpectraMax at 450 nm absorbance. For quantifying Tau in Triton and SDS fractions, cell lysates from sequential extraction were diluted 1000 times in EC buffer. Serially diluted recombinant human T40 was used to generate standard curves. 30 μl of diluted samples were loaded into each well of the plate coated with capture antibody Tau 5. After overnight incubation at 4 °C, captured proteins were reported using biotin-labeled BT2 and HT7, which were left to incubate overnight at 4 °C. The next day, following 1 h of incubation with HRP-conjugated streptavidin at 25 °C, the plate was developed using trimethylbenzidine peroxidase substrate solution.

P301L Mutation Enhances Seeded Recruitment of Endogenous Tau

To assess the efficacy of recruitment by pffs of 4R Tau, with or without N-terminal inserts, as well as disease-associated mutant Tau, fibril transduction was performed on QBI-293 cells transiently transfected with human T43 (0N4R), T40/ΔK280 (FTDP-17 mutation in the MT-binding domain), T40/P301L (FTDP-17 mutation in the MT-binding domain), and T40/R406W (FTDP-17 mutation outside the MT-binding domain). Quantification of the percentage of aggregate-bearing cells induced by these various combinations revealed a similar extent of aggregation in T43- and T40-transfected cells upon Myc-K18 fibril transduction (about 8% of cells had aggregates with either treatment), suggesting the number of N-terminal inserts does not affect fibrillization (Table 1). Among the dominantly inherited Tau mutations in FTDP-17 studied here, T40 with the P301L mutation demonstrated the highest propensity of pff-induced aggregation. Specifically, transduction of Myc-K18 pffs and Myc-K18 pffs with P301L mutation (Myc-K18/P301L fibrils) led to about 20 and 35% of cells bearing aggregates, respectively (Table 1), corresponding to ∼40 and ∼70% of the transfected cells assuming 50% transfection efficiency (Fig. 2A). Although Myc-K18/P301L fibrils appeared to be better “seeds” for T40/P301L recruitment as monitored by number of aggregate-bearing cells, they were comparable with, or even slightly less efficient than, Myc-K18 fibrils in nucleating the fibrillization of WT-T40 (Table 1). Enhanced aggregation of T40/P301L as compared with WT-T40 corresponded to significantly more accumulation of Triton-insoluble Tau as shown on immunoblots (compare Fig. 2B with supplemental Fig. S2B). The induced Tau aggregates were also Sarkosyl-insoluble, with the immunoblot of Sarkosyl extraction indistinguishable from that of Triton extraction (supplemental Fig. S3).

Quantification by Tau sandwich ELISA indicated that Myc-K18/P301L pffs always resulted in higher amounts of Triton-insoluble T40/P301L Tau than Myc-K18 pffs, confirming a more efficient recruitment of soluble Tau with the combination of Myc-K18/P301L pffs and T40/P301L-expressing cells (Fig. 2C). Interestingly, both Myc-K18 and Myc-K18/P301L pff transduction on T40/P301L-expressing cells invariably led to a reduction of Triton-soluble Tau when compared with nontransduced cells, further supporting that the accumulation of insoluble Tau is through sequestration/recruitment of soluble Tau.

Intriguingly, live immunostaining followed by immunofluorescence after fixation and permeabilization (two-stage staining) revealed that the majority of pffs associated with cells were merely membrane-associated despite extensive trypsinization before replating cells to coverslips, and only a small amount of pffs were truly internalized (Fig. 3A). Moreover, double labeling using anti-Myc antibody and PHF-1 in cells extracted with 1% Triton X-100 showed that less than half of the induced aggregates contained detectable colocalizing pffs (Fig. 3B), suggesting that small amounts of pff seeds are sufficient to induce robust endogenous Tau fibrillization.

Open in a separate window

FIGURE 3.

Internalization of small quantities of Tau pffs is sufficient to induce robust Tau aggregation. QBI-293 cells were transiently transfected with mutant T40/P301L Tau and transduced with Myc-K18/P301L pffs using BioPORTER reagent. A, incubation of fibril-transduced cells with polyclonal anti-Myc antibody before fixing (live, red) followed by staining of fixed and permeabilized cells with anti-Myc mAb 9E10 (fixed, green) showed extensive colocalization, suggesting the majority of the pffs detected were merely extracellularly associated with cell membranes. Arrows point to truly internalized fibrils that were only recognized by 9E10 applied to fixed and permeabilized cells but not polyclonal anti-Myc antibody used on live cells. B, double labeling of phospho-aggregates by PHF-1 (green) and exogenous pffs by polyclonal anti-Myc (red) showed very little colocalization. Arrows point to aggregates with colocalizing pff seeds. Aggregates without arrows do not contain obvious pff seeds. Magnification: ×20. Scale bar, 100 μm.

pff-induced Tau Aggregates Are Composed of Filamentous NFT-like Amyloid Structures

Routine EM and immuno-EM using PHF-1 revealed abundant cytoplasmic filaments in pff-transduced cells (Fig. 4) but not in those treated with transduction reagent alone. Moreover, diverse organelles, including mitochondria, lysosomes, and other vesicular structures, were detected within and adjacent to PHF-1-positive filamentous structures (Fig. 4A, panels a and b). The lack of diffuse PHF-1 staining in the cytoplasm outside of the PHF inclusions further supports the concept of seeding and recruitment of soluble Tau to form NFT-like structures in pff-transduced cells.

To examine the progression of pff-induced endogenous Tau aggregation, we fixed cells at different time points after the addition of the pff-reagent complex. At t = 1 h, there was little accumulation of Triton-insoluble Tau, except for a few rare cells showing focal PHF-1 immunoreactivities with small but intensely labeled inclusions surrounded by slender, discrete fibrillar structures (supplemental Fig. S4A). At t = 3 h, focal inclusions became more frequent, and a small population of cells developed skein-like accumulations throughout the cytoplasm (Fig. 5A). It is noteworthy that the earliest focal accumulations were almost always colocalized with Myc immunoreactivities, suggesting that aggregation first begins in cells containing more abundant fibril seeds that rapidly initiate the conversion of endogenous Tau into insoluble aggregates through physical recruitment. At t = 6 h, insoluble Tau was found in about 4% of total cells. Although diffuse fibrillar accumulations remained the most common phenotype at this stage, some aggregates started showing signs of coalescence, and large compact inclusions appeared in a small subset of cells (Fig. 5B). Twenty four hours after fibril transduction, about 25% of total cells were replete with Triton-insoluble Tau, with a significant portion of cells bearing large, densely packed aggregates (Fig. 5C). ThS was used to determine whether pff-induced Tau aggregates form amyloid fibrils. Up to t = 3 h, ThS staining completely overlapped the Myc immunoreactivities, indicating exclusive recognition of β-pleated sheet in Myc-K18/P301L pffs (supplemental Fig. S4B). ThS staining that colocalized with PHF-1-positive endogenous Tau aggregates first emerged at t = 6 h and became more abundant at t = 24 h (Fig. 5, B and C). Interestingly, only the large compact aggregates, but not diffuse skein-like inclusions, were recognized by ThS. We speculate that dispersed fibrillar aggregates probably represent early fibril species that eventually coalesce and mature into β-sheet-rich inclusions.

Newly Synthesized Tau Is Recruited Rapidly into Insoluble Tau Fibrils

To directly monitor the recruitment of newly synthesized Tau into insoluble aggregates, we performed pulse-chase experiments on Myc-K18/P301L pff-transduced T40/P301L-expressing cells. In cells treated with transduction reagent alone, ³⁵S-radiolabeled Tau remained soluble over a 6-h chase period despite a time-dependent electrophoretic mobility shift indicative of phosphorylation of the newly synthesized Tau. In contrast, an increasing proportion of radiolabeled T40/P301L Tau emerged in the Triton-insoluble fraction of fibril-transduced cells, accompanied by more rapid reduction of soluble Tau than in the control cells, demonstrating active recruitment of newly synthesized Tau into an insoluble pool (Fig. 6). After a 1-h chase, an appreciable amount of T40/P301L Tau was already detected in the insoluble fraction, highlighting the rapid solubility change of Tau in the presence of misfolded seeds. Moreover, the detection of slower migrating insoluble radiolabeled Tau suggests continuous recruitment of soluble Tau that is being modified by phosphorylation over time.

Reduced MT Stability Is Associated with Tau Aggregation

Consistent with previous studies in other cell lines, overexpression of Tau in QBI-293 cells led to the formation of thick MT bundles due to excessive stabilization of MTs (21, 36). We hypothesized that seeded recruitment of Tau into insoluble fibrils may attenuate the overstabilization of MTs by sequestering the soluble pool of Tau. As expected, T40/P301L-expressing cells that were transduced with Myc-K18 or Myc-K18/P301L pffs showed significantly less MT bundling revealed by immunostaining of Ac-tub, a marker of stable MTs, as compared with those treated with transduction reagent alone (∼40 and ∼60% reduction in the percentage of cells with MT bundles by Myc-K18 and Myc-K18/P301L pff transduction, respectively) (Fig. 7, A and C). The majority of cells with aggregates showed an absence of MT bundles, and cells that retained bundles typically showed diffuse Tau staining (Fig. 7B). Reduced Ac-tub levels associated with pff-induced aggregation was further confirmed using an Ac-tub sandwich ELISA (Fig. 7D). In addition, there was a significant correlation between Ac-tub level and soluble Tau, but not insoluble Tau, suggesting that the loss of soluble Tau rather then aggregation per se was responsible for the reduced MT stability (Fig. 7, E and F). This observation likely accounts for the occasional MT bundles observed in cells carrying aggregates (data not shown).

Spontaneous Fibril Uptake Is Mediated by Endocytosis

To determine whether Tau pffs can be taken up spontaneously in the absence of protein delivery reagent, we incubated T40/P301L-transfected cells with Myc-K18/P301L pffs for 30 h, followed by trypsinization and incubation for an additional 18 h. Aggregation of endogenous Tau was detected in about 10% of total cells, accompanied by appreciable accumulation of Triton-insoluble Tau detected on immunoblots (Fig. 8, A, panel a, and B). Interestingly, exogenous pffs were barely visible by immunostaining (Fig. 8E), suggesting that internalization of minute quantities of pffs without transduction reagent is sufficient to induce recruitment of endogenous Tau into fibrils. To investigate whether this uptake of Tau pffs occurs by endocytosis, we incubated T40/P301L-expressing cells with Myc-K18/P301L pffs for 4 h at either 37 °C or 4 °C; the latter temperature blocks endocytosis. Because the amount of pff internalized by cells is too low to be easily detected by either immunostaining or immunoblotting (Fig. 8, E and F), we quantified Triton-insoluble endogenous Tau aggregates as an indirect readout of pff entry. Pff incubation at 37 °C for 4 h prior to trypsinization and an additional 44-h incubation resulted in ∼3% of total cells with Tau inclusions, whereas 4 h of pff incubation at 4 °C followed by exactly the same treatments resulted in a dramatic and significant reduction in the percentage of Tau aggregate-bearing cells (Fig. 8, A, C and D), suggesting pff uptake is a temperature-dependent process. The small number of cells with Tau aggregates (∼0.5% of total cells) detected after 4 °C of pff incubation could be due to the internalization of residual membrane-associated fibrils following trypsinization and additional incubation. Moreover, the difference in the extent of aggregation induced by 4 versus 30 h of incubation of pffs (∼3% versus ∼10% aggregate-bearing cells) indicated a time-dependent internalization of fibrils (Fig. 8D). These results suggest spontaneous uptake of pffs probably occurs by endocytosis.