Production of recombinant tau protein

Tau protein was expressed and purified as described before [26] with some modifications. In brief, Escherichia coli BL21 (Rosetta 2 (DE3), Merck) were transformed by 2N4R Tau plasmid (pRK172) and cultured in terrific broth supplemented with ampicillin at 37 °C. Tau expression was induced at OD: 1–1.5 by the addition of 1 mM IPTG. After 6 h of incubation at 37 °C and 180 rpm, cells were harvested by centrifugation at 4000 g for 10 min at 4 °C. The pellet was resuspended in buffer A (20 mM MES pH 6.8, 50 mM NaCl, 1 mM MgSO4, 1 mM EGTA, 1 mM DTT, 1 mM PMSF) and froze at -20 °C overnight. After a freeze-thaw cycle, the suspension was sonicated 3 s/ml (3 s on, 3 s off) with 50% amplitude (UP200St, Hielscher). Streptomycin sulfate (MP Biomedicals) was added up to 1% and the lysate was clarified by centrifugation at 10,000 g for 10 min at 4 °C. The supernatant was collected and NaCl was added to reach 200 mM. Then the extract was boiled at 95 °C for 15 min and incubated on ice for 10 min. The precipitate was removed by centrifugation at 20,000 g for 10 min at 4 °C. The supernatant was collected and dialyzed in buffer A overnight with two times changes. Semi-purified protein loaded into SP HP HiTrap (GE Healthcare) and eluted with a 0–50% gradient of buffer B (20 mM MES pH 6.8, 500 mM NaCl, 1 mM MgSO4, 1 mM EGTA, 1 mM DTT, 1 mM PMSF) by using an ÄKTA™ pure (GE Life Sciences). The fractions were analyzed with Bis-Tris SDS-PAGE 12% and stained with Imperial protein stain (Thermo Fisher Scientific). The pooled fractions were concentrated using 10 kDa Vivaspin 15R (Sartorius) and run into a Superdex 200 Increase 10/300 GL column (GE Healthcare) pre-equilibrated in PBS. The protein was eluted at a 0.3 ml/min rate, and fractions were analyzed with absorbance at 215 nm. The concentration of pooled fractions was determined by the Bicinchoninic acid (BCA) assay kit (Thermo Fisher Scientific), and the recombinant protein was diluted to 6 mg/ml, aliquoted, and preserved at -80 °C following a snap-freezing step in liquid nitrogen.

Fluorescence labeling of Tau Monomers and aggregates

For labeling of Tau Monomers, the recombinant protein at the concentration of 3 mg/ml (~ 65 µM) was labeled with ATTO488-NHS ester or ATTO633-NHS ester fluorescence dye (ATTO-TEC, Siegen, Germany) based on the manufacturer’s instruction. In brief, the protein solution was adjusted to pH 8.3 with a 0.2 M sodium bicarbonate solution (Sigma-Aldrich) and then incubated with 200 µM dye at room temperature for 1 h in the dark. The unbound dye was removed by Bio-Spin 6 size exclusion spin columns (Bio-Rad Laboratories).

For labeling of the Tau aggregates, two approaches were used: (A) The “pre-aggregation labeling” was performed by mixing 10% of pre-labeled monomers (~ 1 label per monomer) with 90% of unlabeled monomers before the incubation in the fibrillization condition. Although the pre-aggregation labeling ensures an equal degree of labeling between monomers and aggregates, this approach reduces the yield of fibril formation, perhaps due to the interference of labeling at residues involved in Tau fibrillization. Thus, a second approach was used to increase the fibril formation yield. (B) The “post-aggregation labeling” was performed by incubating the aggregated mixture after the fibrillization process with 200 µM ATTO dye, similar to the monomers labeling protocol. The excess unbound labels were removed by washing steps during the fractionation process. The labeling was verified by running the protein on SDS-PAGE followed by fluorescence imaging. The degree of labeling (DOL) was calculated by using the below equation:

$$\:\text\text\text=\frac_\times\:\text\text}\text\text\text\text\text\text\right]\:\times\:\:_}$$

Where \(\:_\) is the absorbance of ATTO488 at 500 nm, MW is the molecular weight of 2N4R Tau (= 21.8), \(\:_\) is the extinction coefficient of the dye at its maximum absorbance (ATTO488 = 90,000). The protein concentration of Tau was measured by BCA assay.

The Guanidine Hydrochloride (GuHCl) unfolding assay was performed by treating 3 µg (~ 1 µl) Tau Monomers and aggregates with 2 M GuHCl for 5 min. Unlabeled Tau aggregates were used for ThT assay, and labeled Tau aggregates were used to study the change in ATTO488 fluorescence. The mixture was diluted in ~ 100 µl of ThT assay buffer or PBS before the fluorimetry.

Fibril formation of recombinant tau protein

The fibril formation of 3 mg/ml 2N4R Tau (~ 65 µM) was induced by 130 µM Heparin (~ 3000 kDa, MP Biomedicals) in PBS buffer pH 7.4 supplemented with 1 mM DTT at 37 °C in a 2 ml microtube. Fibrillization of 100–200 µl mixture was accelerated with 1400 rpm shaking in the presence of a 3 mm glass bead using a Thermomixer (Eppendorf Thermomixer). The fibrillization process was monitored by sampling over time, followed by thioflavin T (ThT) fluorescence measurement. For ThT analysis, 2 µl of aggregation sample were added to 98 µl of 10 µM ThT in 10 mM Tris pH 8.0, and fluorescence measurement was done in black 96 well-plate at Ex: 488, Em: 520 by using a CLARIOStar microplate reader (BMG Labtech, Offenburg, Germany). The Finke-Watzky [27] equation was used to fit the normalized ThT data.

The co-factor-free aggregation of 2N4R tau was performed using the previously described protocol [28]. Briefly, 25 µM of protein were aggregated at 37 °C in 25 mM HEPES, 10 mM KCl, 5 mM MgCl2, 3 mM TCEP, 0.01% NaN3, pH 7.2 buffer in a 96 well plate using a Tecan spark plate reader. Three PTFE beads along with double orbital shaking, were used to promote the aggregation. Thioflavin-T (ThT) at a final concentration of 50 µM was used to monitor the aggregation kinetics.

Fibril formation of bovine serum albumin (BSA)

The BSA fibrillization mixture was prepared by dissolving 5 mg/ml BSA in 20 mM Tris pH 7.4. 100–200 µl of the mixture was incubated at 70 °C with 1000 rpm shaking in a 2 ml microtube with a 3 mm glass bead using a Thermomixer (Eppendorf Thermomixer). For making fluorescent BSA aggregates, 4% BSA-CF488A (Biotium) was added to 96% unlabeled BSA before the incubation in the fibrillization condition. The amyloid fibril formation was confirmed by using PROTEOSTAT® Protein aggregation assay (ENZ-51023, Enzo).

Fractionation of tau aggregates

The labeled or unlabeled Tau aggregates obtained from the fibril formation process were divided into four fractions of large fibrils, small fibrils, soluble oligomers and monomers via a stepwise procedure as follows: (I) The fibrillization mixture was transferred into 1.5 ml microtubes with portions of 100–150 µl and centrifuged 16,000 g for 1 h at 4 °C. The pellet was dispersed and washed twice with 100 µl of PBS with the same centrifugation setting and finally dispersed in PBS as large fibrils (L-fib). (II) The supernatant of step one was collected and transferred into 1.5 ml Eppendorf ultracentrifuge tubes with portions of 100 µl and subjected to ultracentrifugation 100,000 g for 30 min at 4 °C. The pellet was washed with 100 µl PBS with the same centrifugation setting and finally was dispersed in PBS as small fibrils (S-fib). (III) The supernatant of the second step was collected and filtered through a 100 kDa ultrafilter (Amicon Ultra 0.5 ml, Millipore) with portions of 100–500 µl for 10 min at 4 °C. The retained phase was washed twice with 500 µl PBS and concentrated as soluble oligomers (Oligo). (IV) The pass-through of step three was washed twice with 500ul PBS using a 10 kDa ultrafilter (Amicon Ultra 0.5 ml, Millipore). Finally, it was concentrated as fibrillization-derived monomers (F-mono). The fractions were aliquoted and stored after a snap freeze in liquid nitrogen at -80 °C.

Native state immunoblot

The samples were collected at different times during the fibrillization process or after fractionation to examine the conformational status of the aggregates using a dot-blot assay. Samples were diluted 1:15 in PBS and loaded into 0.2 μm nitrocellulose blotting membrane (GE, 10600004) using a dot blot chamber (11055, Life technologies). Membranes were washed three times with PBS, then released from the chamber and blocked for 1 h at RT in 30% Roti-Block solution (Carl Roth, Karlsruhe, Germany) before the overnight incubation with primary antibodies at 4 °C under continuous agitation. After 3 times washing in Tris-buffered saline (TBS) supplemented with 0.05% Tween-20 (Sigma-Aldrich) (TBST), HRP-coupled secondary antibodies were incubated for 1 h at room temperature. After another round of washing steps in TBST, for visualization, membranes were incubated in Clarity Western ECL Substrate (Bio-Rad Laboratories), and imaging was done with Odyssey Fc (LI-COR Biotechnology, Lincoln, NE) imaging system. The primary antibodies used in this study: Tau-5 (1:1000; MAB361, Merck Millipore, Billerica, MA), a general monoclonal anti-Tau antibody; TNT-1 (1:1000; MABN471, Merck Millipore, Billerica, MA) identifying the phosphatase-activating domain in the N-terminal region of Tau, which is only exposed in a pathological conformation [29]; TOMA-1 (1:500; MABN819, Merck Millipore, Billerica, MA) being an anti-Tau oligomer-specific conformational antibody [30]; MC1 (1:500; a gift from Dr. Peter Davis) indicating a pathological conformation by binding two discontinues AD-specific epitopes at N-terminal and microtubule-binding domain [31]. The anti-mouse IgG (1:2000; Vector Laboratories, Burlingame, CA) was used as a secondary antibody.

Biophysical characterization of tau species

Size exclusion chromatography was performed by pre-equilibration of Superdex 200 Increase 10/300 GL (GE Healthcare) with two column volumes of elution buffer (PBS) followed by loading 500 µl of samples at 0.1 mg/ml concentration. Samples were injected into an ÄKTA™ pure (GE Life Sciences) and run at flow rates of 0.3 mL/minute. The elution profile was monitored at wavelengths of 214 and 280 nm. The fractions were collected and further analyzed using dot-blot assay and dynamic light scattering (DLS). DLS measurements were carried out using a Malvern Zetasizer-Nano instrument with a 4 mW He-Ne laser (633 nm) in a water suspension at 0.03 mg/ml concentration.

Atomic force microscopy (AFM) imaging was performed using the NanoWizard® 4 (JPK, Berlin, Germany) and SPM software with an integrated Axiovert 200 inverted microscope (Zeiss, Jena, Germany). The cantilevers qp-BioAC-CB1 (NanoWorld, Neuenburg, Switzerland) with a resonance frequency of 90 kHz and a spring constant of 0.3 N/m were used and calibrated with the contact-free method. The QI™ Mode (Advanced Imaging) and the following parameter settings were used for the image acquisition: setpoint 0.4–0.6 nN; z-length 86–126 nm and pixel time 2.2–5.5 ms. The sample was diluted in distilled water and dried on a freshly prepared surface of the highest grade V1 AFM Mica Discs, 10 mm (Ted Pella). Mica discs were washed three times with distilled water. The measurements were performed in air at ambient temperature. All images were processed, optimized, and zoomed in with the data processing software version 6.0.50 (JPK, Berlin, Germany). The first step was subtracting a polynomial fit from each scan line independently. A histogram was calculated for each scan line, and the data between the lower (0%) and upper (70%) limits was used to fit the polynomial. The next step was replacing outlier pixels with the median value of neighboring pixels. Lastly, a low-pass filter was applied (2-dimensional Savitzky–Golay smoothing; smoothing width: 5, order: 4).

Transmission electron microscopy (TEM) of Tau aggregates was performed as described before [32]. In brief, 3 µl of Tau aggregates were loaded onto glow-discharged 400 mesh Formvar/carbon grids (EM resolutions) for 20 s, blot-dried, and stained three times with uranyl formate (3 µl, 15 s for each time). TEM imaging was carried out by using a Tecnai G2 Spirit BioTWIN (FEI) operating at 120 kV acceleration. Images were obtained on a TemCam-F416(R) (TVIPS) CMOS camera.

UV-Circular dichroism (UV-CD) spectroscopy of different fractions of Tau aggregates was performed via a Chirascan V100 CD spectrometer (Applied Photophysics) by loading 2–6 µg of protein samples. The UV-CD spectra were recorded between 190 and 250 nm with a step size of 1 nm and a scanning speed of 10 nm/min using a 1-mm path-length cuvette at room temperature.

Gradient centrifugation of iodixanol was performed by manual filling of 10 ml gradient columns with ten portions of 1 ml OptiPrep of 5–50% (Sigma, D1556) in 14 ml centrifuge tubes (Beckmann Coulter, 344060). A 500 µl of each fraction of ATTO-488 labeled Tau, including L-fib, S-fib, and soluble fraction (the supernatant of 30 min centrifugation at 100.000 g at 4 °C), were loaded on top, and then columns were subjected to ultracentrifugation 250,000 g at 4 °C for 3 h. The 10 ml columns were fractionated manually into 40 fractions of 250 µl in black 96-well plates, and the fluorescence of each fraction was measured at Ex: 488, Em: 535 by using a CLARIOStar microplate reader (BMG Labtech, Offenburg, Germany).

Induced pluripotent stem cells derived neurons (iPSCNs) culture and differentiation

For the ease of differentiation via lentiviral transduction, small molecule neuroprogenitor cells (smNPCs) were generated from induced pluripotent stem cells via embryoid body formation and stable integration of an inducible NGN2 vector, as described before [33, 34]. NGN2_smNPCs were cultured in N2B27 medium (48.425% DMEM/F12 Medium, 48% Neurobasal Medium, 0.5% N2-supplement, 1% B27 supplement without Vitamin A (Life Technologies, Carlsbad, CA, United States), 0.025% Insulin (Sigma-Aldrich, St. Louis, MO, United States), 0.5% Non-essential amino acids, 0.5% GlutaMax, 1% Penicillin/Streptavidin (Life Technologies, Carlsbad, CA, United States) and 0.05% β-Mercaptoethanol) supplemented with 0.5 µM Purmorphamine, 3 µM CHIR99021 and 64 µg/ml ascorbic acid (Th. Geyer, Renningen, Germany). During expansion and differentiation, cells were maintained at 37 ‌°C with 5% CO2 in a tissue culture incubator.

For the neuronal differentiation, plates were coated by overnight incubation at 4 °C with 100 µg/ml Poly-L-ornithine (PLO) (Sigma-Aldrich, St. Louis, MO, United States) diluted in 0.1 M borate buffer at pH 8.4 and subsequent incubation with 10 µg/ml Laminin (Sigma-Aldrich, St. Louis, MO, United States) at 37 °C. NGN2_smNPCs were directly seeded in an induction medium containing N2B27 medium supplemented with 2.5 µg/ml doxycycline (Sigma-Aldrich, St. Louis, MO, United States) and 2 µM DAPT (Cayman, Ellsworth, United States). The medium was entirely changed at day 3 to N2B27 supplemented with 2.5 µg/ml doxycycline, 10 µM DAPT, 10 ng/ml BDNF, 10 ng/ml GDNF, 10 ng/ml NT 3 (PeproTech, Princeton, NJ, United States) and 0.5 µg/ml laminin. From day 6, only 50% of the medium was changed every 3 days with fresh differentiation medium without doxycycline and DAPT.

Lund human mesencephalic (LUHMES) cell culture and differentiation

As described before [35], LUHMES were cultured in flasks (EasYFlasks, Nunclon DELTA, VWR, Darmstadt, Germany) coated with 50 µg/ml PLO (Sigma-Aldrich, St. Louis, MO) in expanding medium containing DMEM-F12 (Sigma-Aldrich) supplemented with 1% N2 supplement (Life Technologies, Carlsbad, CA) and 0.04 µg/ml basic fibroblast growth factor (bFGF; PeproTech, Rocky Hill, CT). For differentiation, plates were coated first with 50 µg/ml PLO (Sigma-Aldrich) and then with 5 µg/ml bovine fibronectin (Sigma-Aldrich). Cells were seeded directly in the differentiation medium containing DMEM/F12 with 1% N2 supplement, 1 µg/ml tetracycline, 0.5 µg/ml dibutyryl cyclic-AMP (Sigma-Aldrich), and 2 ng/ml glial cell-derived neurotrophic factor (GDNF; R&D Systems, Minneapolis, MN). During expansion and differentiation, cells were maintained at 37 °C with 5% CO2 in a tissue culture incubator.

Cell viability assay

For studying the toxicity of Tau species, smNPCs were cultures for differentiation in transparent bottom black 96 well-plates (PerkinElmer) and, at day 13–15th of differentiation, treated with 250 nM (Monomer equivalent) of Tau species for 24 h. Then media was removed and viability was assessed using the cell viability indicator of the neural outgrowth staining kit based on the company instruction (A15001, life technologies). The fluorescence was measured by CLARIOStar microplate reader (BMG Labtech, Offenburg, Germany) with a matrix of 15 × 15 from the bottom (Ex: 480 nm, Em: 520–535 nm).

Tau endogenous aggregation biosensor assay

Seeding in HEK293T overexpressing mutant P301S 0N4R Tau, C-terminally tagged with Venus protein, was performed as described before [36]. In brief, cells were cultured in complete DMEM (C-DMEM) with 10% (vol/vol) FCS, 100 U/ml penicillin, and 100 µg/ml streptomycin at 37 °C in a 5% CO2 atmosphere. Cells were cultured on poly-L-lysine (Sigma, P4707) coated transparent bottom black 96-well plates in C-DMEM (PERK6055302, PerkinElmer) for seeding. On day 2, the media were discarded and the wells were washed twice with PBS. Seeding was induced by adding OptiMEM (Gibco™, 51985026) containing 1% (vol/vol) Lipofectamine 2000 (Life Technologies) and 400 nM (Monomer equivalent) of Tau species for 1 h. Next, the seeding medium was aspirated and C-DMEM was replaced. On day 4, the media were changed to FLuoroBrite DMEM (Gibco™, A1896701) containing 1X backdrop suppressor (Thermo Fischer Scientific, B10512), and cells were imaged by fluorescence microscope and plate reader.

Neuronal entry assay

A live-cell neuronal entry assay was carried out, as described before [37], by using a split luciferase called NanoLuc (Nluc) composed of a large 18 kDa subunit (LgBiT) and a small 11 amino acid peptide (HiBit) forms a complementation reporter [38]. HiBiT-tagged Tau was added to the extracellular medium of primary mouse neurons expressing LgBiT. Tuck et al. previously showed that human and mouse neurons had similar Tau uptake dependencies [37]. To do this, using the abovementioned method, 0N4R P301S-Tau-HiBiT were fibrillized and fractionated into four different fractions of L-fib, S-fib, Oligo, and F-mono. Primary neurons were prepared from postnatal day 0/1 C57BL/6 mouse pups and infected at 2 days in vitro (DIV) with AAV1/2 hSyn::-eGFP-P2A-LgBiT-nls particles at a multiplicity of 50,000 genome copies per cell. On DIV 7, 2 ug/ml of each Tau-HiBiT species was prepared in maintenance media. 50% of the media was used for an in vitro reconstitution assay, and the remainder was used for neuronal entry assays. For in vitro reconstitution, the total signal in the maintenance media (RLU in media) was quantified by the addition of excess recombinant LgBiT for 30 min. With the remaining media, neurons were 100% media changed and incubated with Tau-HiBiT preparations for the depicted amount of time. Cytosolic entry was quantified (RLU in cells), followed by incubation for 42 min with PrestoBlue cell viability reagent according to manufacturer instructions (Thermo Fisher Scientific). Fluorescence intensity was quantified (excitation 540–570 nm; emission 580–610 nm) using the CLARIOstar microplate reader (BMG Labtech). Total viable cells per well were calculated using a standard curve of viable cells per well and adjusted for background fluorescence. Percent cytosolic entry normalized to cells was calculated by dividing the RLU in cells by RLU in media and normalizing it to total viable cells per well.

Uptake and accumulation assay

smNPCs or LUHMES were seeded in a black clear-bottom 96-well tissue culture plate (PERK6055302, PerkinElmer) for differentiation. Neurons were treated with fluorescently labeled Tau species in differentiation media after complete media removal. The concentration and incubation time varied for different experiments and were specified in the result section. The FL-Tau-containing media were removed and 100 µl FluoroBrite DMEM (Gibco™, A1896701) containing 1X BackDrop background suppressor (B10512, Thermo Fisher Scientific) was added. The fluorescence of ATTO488 was scanned by excitation at 490 nm and emission at 510–530 nm (Focal length: 0.9 mm, Gain: 2200) with a matrix of 15 × 15 from the bottom of the wells by using a CLARIOStar microplate reader (BMG Labtech, Offenburg, Germany). The fluorescence intensity of each well was normalized by dividing to the untreated cells as a blank for background fluorescence. The representative fluorescence images were taken by live imaging using Leica DMI6000 B (Leica Microsystems, Germany). To validate the comparability of cellular uptake and accumulation under various treatments and conditions, cells’ viability was examined using the 0.1 μm Calcein-AM (Thermo, C3100MP) treatment for 30 min, and measurement was performed in the presence of 1X BackDrop background suppressor (B10512, Thermo Fisher Scientific). The treatment conditions were adjusted in the none-toxic range.

Recombinant 2N4R Tau Monomers were labeled with ATTO488 to compare the uptake of monomer and aggregated mixtures. Then, the labeled monomers were mixed 1:9 with unlabeled monomers. Next, the mixture was divided into two parts; part A was kept at 4 °C, and part B was incubated in the fibrillization condition described as “pre-aggregation labeling” in the fibril formation section. iPSCNs at day 13–18 of differentiation were treated with part A (Mono) and part B (Agg) at various concentrations and different incubation times (at 37 °C) to compare the kinetics and titration. For comparing the uptake of Tau fractions, “post-aggregation labeling” was used as described in the fibril formation section. iPSCNs at day 13–18 of differentiation and LUHMES at day 6–8 of differentiation were treated with labeled fractions at various concentrations and incubation times to compare the kinetics and titration.

For the Tau competition assay, cells treated with 50 or 100 nM (Monomer equivalent) of ATTO488 labeled Tau Mono or aggregates and at the same time with 4- or 5-times higher concentrations of unlabeled species for 16–20 h. Since the uptake of Mono was lower than aggregates, a 10-fold higher degree of labeling (labeling efficiency around 2 compared to ~ 0.1–0.2 for aggregates) was used. It is crucial that competing aggregates are from the same batch of aggregate preparation.

To study the impact of the small molecules on Tau uptake, iPSCNs were treated at days 13–18 of differentiation. For the small molecules in Table 1A, cells were treated for 30 min at a specified concentration. After a washing step (with 100 ul PBS), they were incubated with 250 nM Tau for 3 h. For the small molecules in Table 1B, cells were co-treated with small molecules at a specified concentration and Tau at 25–50 nM for 18–20 h. For viability assessments, cells were treated with a standard medium for 3 h and then incubated for 30 min with 0.1 μm Calcein-AM (Thermo, C3100MP).

Table 1 List of small molecules used for studying the intracellular accumulation of eTau

For the pre-treatment experiments, LUHMES were treated with 100 µM Heparin, 200 nM monomers, or small fibrils (Monomer equivalent) for two hours on day six. Then, the media was removed, and cells were washed once with PBS before treatment with fluorescently labeled Tau Monomers or small fibrils.

For the gene-knockdown experiments, cells were treated with 10 nM siRNAs (LRP1, EXT1, and EXT2: siPOOLs from siTOOLs, VPS35: Silencer Select siRNAs from Thermo Fischer) in the presence of 0.075 µl/well RNAiMax lipofectamine (Thermo Fisher Scientific, 13778075) in 1:1 differentiation medium to OptiMEM (Gibco™, 51985026) for 24 h. siRNA treatments were performed on day 2 post-seeding, and the media was changed the next day. iPSCNs and LUHMES neurons were treated with 25–50 nM (Monomer equivalent) of Tau Monomers and small fibrils for 18–20 h on days 10–12 and 6–8 of differentiation, respectively. The knockdown was confirmed by immunoassay.

Western blot

iPSC-derived neurons and LUHMES neurons were differentiated in 6-well plates and treated with siRNA as described before. The whole cell extract was collected by radioimmunoprecipitation assay (RIPA) lysis buffer (Thermo, 89901) as described by the manufacturer. For western blot, 20–40 ug of cell extracts were loaded in 4–12% Bolt Bis-Tris precast Gels (Invitrogen, NW04127BOX) and were run for 20 min by a PowerEase Touch 350 W (Invitrogen, PSC350M), then were transferred to methanol-activated low fluorescence 0.2 μm PVDF membrane (GE, 10600022) by a PowerBlotter XL (Invitrogen, PB0013) at 25 V, 2.5 A for 13 min. The membrane was blocked with 1X Roti-Block (Carl Roth, A151.2) for 1 h at RT and incubated overnight at 4 °C with primary antibodies in 1X Roti-Block diluted in wash buffer (0.05% Tween-20 in Tris-buffer saline (TBS-T) pH 7.5). Following 3 times washing in TBS-T, the membranes were incubated with a secondary antibody in 1X Roti-block in TBS-T for 2 h at RT. After three times rinsing with TBS-T, the membranes were incubated 5 min at RT with SuperSignal West Pico PLUS Chemiluminescent Substrate (Thermo, 34580) and imaging was performed by iBright CL1500 imaging system (Invitrogen, A44114). Loading control was performed by incubation of the membranes with a β-actin (1:1000, Cell Signalling Technology 13E5) and the following primary antibodies were used in this study: LRP1 (1:1000; Abcam ab92544), EXT2 (1:100, Santa Cruz Biotechnology sc514092), VPS35 (1:1000, Cell Signalling Technology E6S4I). The following secondary antibodies were used: HRP-coupled anti-mouse (1:2000; Cell Signalling Technology 7076), or -rabbit antibody (1:5000; Cell Signalling Technology 7074).

Confocal imaging

smNPCs and LUHMES were plated on 8-well ibidi µ-slides (ibidi, Gräfelfing, Germany) and incubated in the differentiation medium for 15 and 7 days, respectively. Following the differentiation, neurons were treated with Tau Mono labeled with ATTO633 and Tau S-fib labeled with ATTO488 for 3 h. Imaging was done immediately or 21 h after treatment. In the latter case, cells were washed and incubated in the differentiation medium until imaging. Before imaging, cells were stained with 1 µM Cell Trace™ Calcein Red-Orange (Thermo Fisher Scientific) for 30 min. Then the media were changed to FluoroBrite DMEM containing 1X Backdrop™ background suppressor (Thermo Fisher Scientific). Live Z-stack images were taken using Zeiss LSM 880 (Carl Zeiss, Oberkochen, Germany) via a 63x oil immersion objective and 2X digital zoom.

Statistical analysis

Statistical analysis was performed using GraphPad Prism 8.0.2 (GraphPad Software, La Jolla, CA, USA) or the Excel data analysis package. All data shown in the figures are presented as mean ± standard error of the mean (SEM) or standard deviation (SD). All data were analyzed via one-way ANOVA followed by Dunnett’s post hoc test, except for the siRNA experiment, where two-way ANOVA followed by the Sidak test was performed. P-values < 0.05 were considered statistically significant.