Halting of Caspase Activity Protects Tau from MC1-Conformational Change and Aggregation

Materials

The 2N4R Tau isoform was cloned into pRc/CMV2. The tau-specific CP27 (aa 130–150), DA9 (aa 102–130), TG5 (aa 220–235), and MC1 (aa 7–9 and aa 312–342) antibodies were generated and characterized as described [11, 35]. Staurosporine, pan-caspase inhibitors: Z-VAD(Ome)-FMK and ApoBlock were purchased from Calbiochem, Axxora, and BD Biosciences, respectively. Anti-Tau 421 antibody (tau-C3) was from Abcam, and anti-active caspase-3 and anti-active caspase-6 were purchased from Abcam and Sigma, respectively.

Cell culture

The murine neuroblastoma Neuro-2a (N2a) cell line was maintained in 50% DMEM and 50% Optimem supplemented with 5% fetal bovine serum (FBS), 1% penicillin/streptomycin and 1% MEM essential amino acids, at 37°C with 5% CO₂. N2a cells were plated in standard 6-well plates for transient transfections.

Transient transfections with 2N4R tau isoform were performed using Lipofectamine 2000 (Invitrogen) with a total amount of 0.25–2μg cDNA per 1.6 cm² surface area in 6-well plates, or 15μg cDNA per T75 flask, at a DNA:Lipofectamine ratio of 1:3.

To induce cellular stress and caspase induction, at 24 h after transient transfections cells were treated with 0.0325–1μM staurosporine for 6–24 h. Pan-caspase inhibitors were used at 50μM concentrations with incubation times of 6–24 h.

Purification of paired helical filaments

PHFs were purified from AD brain tissue as described in Jicha et al. [14]. Briefly, AD cortical tissue was weighed and homogenized in TBS buffer (containing 1 mM PMSF) and 1×complete protease inhibitor cocktail (Roche Diagnostics) at a ratio of 10 ml buffer per g tissue. The homogenate was spun at 28,000× g for 30 min at 4°C. The supernatant was reserved and the pellet was discarded. The supernatant was then run over a 25 ml MC1-Affigel 10 column, with a 4 cm high guard column of Sepharose 400 Superflow, with a flow rate of 50–60 ml/h. The supernatants were recycled twice through the column over 18–20 h at 4°C. The bound PHF was then eluted with KSCN, and the protein containing eluate was pooled and dialyzed overnight. The buffer was exchanged to TBS. An AT8 ELISA and western blot was then run to determine the PHF concentration in the sample.

Preparation of cell lysates

After removing the medium, cells were rinsed with Tris-buffered saline solution (TBS) and briefly sonicated in homogenization buffer (10 mM NaF, 1 mM Na₃VO₄, 2 mM EGTA, in TBS, pH 7.4) supplemented with Complete Mini Protease Inhibitor Cocktail tablets (Roche Diagnostics). Cell lysates were cleared by centrifugation at 2600× g for 10 min at 4°C. Supernatants were collected and protein concentration was estimated using the BCA protein assay kit (Fisher Scientific UK). Cell lysates were stored at –80°C.

Preparation of brain tissue homogenates

Human AD and control brain samples (AECOM collection) were sonicated in homogenization buffer (50 mM Na₃PO₄, pH 7.0, 10 mM Na₄P₂O₇, 0.5 mM PMSF, 2 mM Na₃VO₄, 2 mM EGTA, 2 mM EDTA, 20 mM NaF, 1 mM DTT supplemented with Complete Mini Protease Inhibitor Cocktail tablets; Roche Diagnostics) at a ratio of 10 ml buffer per gram of tissue. Homogenates were cleared by centrifugation at 13 000× g for 5 min at 4°C. Supernatants were collected and protein concentration was estimated using the BCA protein assay kit (Fisher Scientific UK). Cell lysates were stored at –80°C.

GUAVA ViaCount and MultiCaspase assays

Cell viability and caspase induction were measured in a GUAVA Technologies Inc PCA-96 flow cytometer using ViaCount and MultiCaspase assays (Millipore) according to the manufacturer’s protocols. The analyses were done using GUAVA software that allowed selection of four population of cells: i) negative for both dyes, ii) positive for SR-VAD-FMK and negative for 7-AAD, iii) cells positive for both dyes, and iv) only 7-AAD positive to be segregated into separate quadrants, which reported the percent of total, and mean intensities are for each quadrant.

Lactate dehydrogenase (LDH) release assay

N2a cell medium was subjected to the CyTox 96^® Non-Radioactive cytotoxicity assay. The positive control used was a sample of lysed cells that was harvested in the provided proprietary lysis buffer. Once all samples had been harvested and the positive control generated, 50μl of each sample was transferred to a flat-bottomed 96-well plate, and 50μl of the kit substrate was added to each well. The plate was incubated in the dark for 30 min, before addition of 50μl of stop buffer. The absorbance at 490 nm was recorded, and data were plotted as a percentage of the positive control.

ELISA assays

Flat-bottomed 96-well plates were coated with either DA9 (1μg/ml) or MC1 (2μg/ml) antibodies in coating buffer (15 mM K₂HPO₄, 25 mM KH₂PO₄, 0.8% NaCl, 1.2 mM EDTA, 0.05% NaN₃, pH 7.2) and incubated overnight at 4°C. The following day the plates were washed in wash buffer (100 mM NaCl, 10 mM Tris base, 1% Tween 20, pH 7.2) and blocked with Starting Block (Pierce). Prior to serial dilutions the total protein concentration in cell lysates and brain homogenates was adjusted to 0.5 mg/ml. Serial dilutions of cell lysates or brain homogenates were prepared in TBS with 20% Superblock (Pierce). Serial dilutions of the control experimental samples or PHFs were used as standard curves. The following day the plates were washed prior to being incubated with CP27 (1μg/ml) antibodies followed by incubation with the secondary goat anti-mouse IgG2b-HRP antibody (1:2000). 1-Step Ultra-TMB (Pierce) was used to develop the reactions and 2 M sulphuric acid to quench them. The results were read using absorbance measurements at 450 nm.

Fractionation of cell lysates

Fractionation of cell lysates was performed as previously described [36], with minor modifications. To obtain the low speed (LS) fraction, cell lysates underwent a 10-min centrifugation at 2600× g at 4°C. Part of the LS fraction was mixed with an equal volume of lysis buffer (20 mM Tris-HCl, pH 7.4, 140 nM NaCl, 1 mM phenylmethylsulfonyl fluoride, 1 mM Na₃VO₄, 1 mM EDTA, 1X complete protease inhibitors) supplemented with 1% v/v Nonidet P40, and centrifuged for 1 h at 100,000× g at 4°C. The supernatant was collected and stored as the S1 fraction. The pellets (P1) were washed three times with lysis buffer supplemented with 0.5% v/v Nonidet P40. Between each wash the samples were centrifuged at 15 000× g for 3 min at 4°C. For ELISA assays, the pellets were resuspended in a third of the initial LS sample volume of lysis buffer. All samples were stored at –80°C. For western blotting, the pellets were solubilized in 8 M urea by gentle agitation at 37°C before being resuspended in Laemmli buffer (0.125 M Tris-HCl pH 6.8, 5% w/v sodium dodecyl sulphate, 10% w/v 2-mercapto-ethanol, 20% v/v glycerol and enough bromophenol blue) and heated for 5 min at 90°C. Samples were stored at –20°C.

Aggresome detection

N2a cells were prepared for aggresome detection using the FACS Canto and confocal microscopy. For confocal microscopy analysis of aggresomes, N2a cells plated in 96-well plates (BD Biocoat) were transfected with tau and treated with staurosporine, before fixation with Glyo-Fixx (ThermoFisher Scientific) containing 0.1% Triton X-100 (Sigma Aldrich) for 20 min at room temperature. Cells were blocked with 5% marvel milk powder in PBS for 1 h before incubation with the anti-CP27 antibody (1:1000) overnight. The following day cells were washed 3 times in PBS before incubation with the secondary antibody Alexa-Flour IgG1 488 (1:1000) (Life Technologies), aggresome detection reagent (1:2000) (Abcam) and Hoechst dye (1:1000) (Life Technologies). Cells were incubated for 1 h at room temperature, before imaging. For analysis of aggresome formation using FACS Canto, N2a cells plated in T75 flasks were transfected with 2N4R tau or were treated with the positive control for aggresome formation (MG-132). Cells were fixed and permeabilized using the Fix + Perm^® Cell Permeabilization Kit (Invitrogen), before incubation with the aggresome detection reagent (1:20,000) for 25 min. Cells were then assessed for aggresome formation immediately by FACS.

Western blotting

Samples were loaded on a NUPAGE RTM Novex 4–12% Bis-Tris Gel (Invitrogen) and run in 1X MES running buffer (Invitrogen) before being transferred onto the Amersham Hybond ECL nitrocellulose membrane (GE Healthcare). Membranes were blocked for 1 h in blocking buffer (5% w/v dried skimmed milk in PBS buffer with 0.05% Tween 20) prior to incubation with primary antibodies in blocking buffer. After incubation with secondary antibodies tau was detected using the ECL Plus Western Blotting Detection Reagents (Fisher Scientific) on the Amersham Typhoon 9400 and band intensities were quantified using the ImageQuantTL program.

Fluorescence-activated cell sorting (FACS)

N2a cells were cultured in T75 flasks, transiently transfected with 2N4R tau cDNA, with or without 0.158μM staurosporine treatment. After trypsinization, cell suspensions were collected and centrifuged at 478× g for 5 min at 18°C before being resuspended in FACS buffer (0.1 % NaN₃, 5 % FBS in PBS). The cell count was adjusted to 1×10⁷ cells per milliliter. Cells were fixed and permeabilized (reagent A and B, Fix + Perm^® Cell Permeabilization Kit, Invitrogen) before immunostaining with anti-active caspase-3, anti-active caspase-6, and and IgG controls, at 10μg/ml. After a 20-min incubation at room temperature and centrifugation at 500× g for 5 min at 4°C, the cell pellets were resuspended in 200μl FACS buffer and the following secondary antibodies were added: anti-rabbit IgG-APC 1:50 (Prozyme) and anti-mouse IgG₁-AF488 1:1250 (Invitrogen). Following a 20-min incubation at room temperature, 3 ml of FACS buffer were added to each tube and centrifuged at 500× g for 5 min at 4°C. The resulting cell pellets were resuspended in 2 ml of FACS buffer before measurement in BD FACS Cantotrademark.

Confocal microscopy

N2a cells were cultured on glass-bottomed culture dishes (MatTek). Following transient transfection with 2N4R tau and 24-h treatment with staurosporine, when required, cells were washed with DPBS (Gibco, Invitrogen), fixed with 4% paraformaldehyde for 10 min and then permeabilized with 0.1% Triton X-100 (Sigma) for 15 min. Alternatively N2a cells were cultured in 96-well plates (BD Biocoat) and fixed with Glyo-Fixx (ThermoFisher Scientific) containing 0.1% Triton X-100 for 20 min. Cells were then washed 3 times with PBS before incubation with blocking buffer (PBS + 5% milk (Marvel)) for 1 h followed by an overnight incubation at 4°C with the primary antibodies MC1 (1:500) or CP27 (1:1000), which were diluted in PBS + 5% milk. Cells were then washed 3 times in DPBS before addition of the secondary antibodies: Alexa Fluor 488 Goat Anti-mouse IgG2b, Alexa Fluor 647 Goat Anti-Mouse IgG1 or Alexa Fluor 568 Goat Anti-Mouse IgG₁, Alexa Fluor 350 and Goat Anti-Mouse IgG_2B, along with Hoechst dye (1:1000) or the aggresome detection reagent (1:2000) when applicable for 1 h at room temperature. The cells were then washed 3 times with PBS before imaging using the confocal microscope (Olympus Fluoview FX), with a 20× or 40× objective and either a 2 or 4 times zoom. Fluorophores were excited with 405 nm, 488 nm, and 633 nm wavelength lasers.

AlphaScreen assays

AlphaScreen assays (Perkin Elmer) were performed according to the manufacturer’s guidelinesusing tau-specific antibodies. Optimised Alpha-Screen assays were performed using either biotinylated DA9 (bDA9) and acceptor (Ab-ACC) bead antibodies: TG5 total tau, CP13 pS202-tau, MC6 pSer235-tau, PHF1 pSer396/pSer404-tau, or anti-Tau 421, and 10μl per well of optimized antibody mix of Ab-ACC and biotinylated antibody were added to 384-well assay plates (Greiner) together with 5μl per well of sample or standard diluted in AlphaScreen assay buffer (0.1% casein in DPBS). Plates were incubated overnight in the dark at 4°C. After overnight incubation, 5μl of the streptavidin-coated donor beads diluted in AlphaScreen buffer were added to each well and plates incubated in the dark with gentle agitation at room temperature for 4 h. Plates were read at an excitation wavelength 680 nm and emission at 520–620 nm using an Envision plate reader (Perkin Elmer). Linear regression with variable slope analyses was used to calculate relative amounts of total tau and phospho-tau from standard curves derived using purified PHFs. Relative amounts of D421-caspase cleaved tau were calculated from the AlphaScreen count levels relative to the standard curves derived using N2a cells transiently expressing 2N4R tau and treated with 0.5μM staurosporine.

Early conformational changes of tau in a cell-based model can be followed in an ELISA assay with MC1 conformational antibody

To establish the relationship between tau levels and generation of MC1-reactive conformational epitopeN2a cells were transiently transfected with the human 2N4R tau-expressing plasmid and analyzed by a newly-established sandwich ELISA using two antibodies for coating, pan-tau-specific DA9 and human tau-specific CP27, and MC1 for detection of conformationally-altered tau. AD and control brain homogenates were used as controls, the former yielding a high signal and the latter a low background signal (data not shown).

Transient transfections of N2a cells with increasing amounts of the 2N4R tau construct led to a rise in total tau expression and concomitant increase in the amount of conformationally-changed tau (Fig. 1A, B). As shown in Fig. 1C, there is an excellent correlation between the relative amounts of total tau and conformationally-changed tau in N2a cells expressing human tau. However, it is noteworthy that not all of the cells expressing high levels of tau are MC1 positive, as shown in the immunofluorescence experiment with MC1 and CP27 double staining (Fig. 1D–F). This suggests that although the amount of conformationally changed tau produced is correlated with the level of tau expression, at the individual cell level there are perhaps additional factors that determine whether MC1-tau is generated.

Tau-expressing N2a cells show a signature of cellular stress

To further characterize our cellular model, we have evaluated the effects of tau overexpression on cell responses (Fig. 2A–E). Overexpression of tau for 48 h led to an increase in membrane permeability, as measured by LDH release, an indication of cellular toxicity (Fig. 2A, B). There was a 30% increase in membrane permeability related to tau expression for 48 h compared to the untransfected controls at the same time point or tau expression for 24 h. Importantly this was accompanied by an increase in the level of caspase-cleaved tau at Asp421, a known cleavage site for caspase-3 and caspase-6 [37–39], which shows that the overexpression of tau causes cellular stress manifested by activation of caspases (Fig. 2C). There was an apparent elevation of phosphorylated tau at Ser202, however, this did not reach statistical significance, with no change in the levels of phosphorylated tau at pSer396. The above observations confirm that tau expression itself causes a cellular response that is likely to participate in the observed generation of conformationally-changed tau recognized by the MC1 antibody.

In N2a cells expressing human tau, treatment with staurosporine induces expression of caspases and concomitant cleavage of tau at Asp421

There is some evidence that the proteolytic processing of tau may play a crucial role in the aberrant change of tau conformation and its aggregation, or is at least an important marker of the aggregation process. The activation of caspases has been implicated in the formation of neurofibrillary tangles in AD [26]. These observations prompted us to determine whether in our experimental system treatment of cells with staurosporine would drive caspase activation and tau aggregation. Neuroblastoma cells were transiently transfected with 2N4R human tau cDNA and then treated in a dose-dependent manner for 24 h with staurosporine. General caspase activation and apoptotic events were followed in a flow cytometer using a biochemical assay for detection of activated caspases (Fig. 3A). This method uses a fluorochrome-conjugated inhibitor of caspases (SR-VAD-FMK) that binds to multiple caspases, and the resulting fluorescent signal is proportional to total cellular caspase activity. Additionally membrane structural integrity was detected by the cell impermeable dye 7-AAD. The assay allowed us to identify i) cells with low levels of caspase activation, which are negative for both dyes, ii) cells with moderate levels of caspase activation being positive for SR-VAD-FMK and negative for 7-AAD, iii) cells positive for both dyes, which have high levels of caspase activation and go through the late stages of apoptosis, and iv) dead cells, which are only 7-AAD positive. Treatment with staurosporine resulted in a dose-dependent decrease of the viable cell population with a concomitant increase of cells with high levels of caspase activation. Even the lowest concentration (0.0625μM) of staurosporine effectively triggered caspase activation, where 29.5% of SR-VAD-FMK-positive cells, and 45% of SR-VAD-FMK and 7-AAD-positive cells of the total cell population had moderate and high levels of caspase activation, respectively. The highest levels of caspase induction were observed after treatment of cells with 0.5–1μM staurosporine, and under these conditions more than 80% of cells had high levels of caspase activation 24 h after treatment (Fig. 3A). An additional follow-up measurement of membrane permeability manifested by LDH leakage under the above experimental conditions showed no differences in membrane permeability between treatment conditions over 24-h treatment with a range of staurosporine concentrations (Supplementary Figure 1A). However the largest increase in membrane leakage was caused by tau expression itself as already shown in Fig. 2B.

Caspase-3 and caspase-6 have been previously indicated to play roles in the cleavage of tau and association with tau pathology in AD [28–31, 40]. Induction of caspase-3 and caspase-6 in our experimental system was confirmed in fluorescence-activated cell sorting (FACS) experiments where over a 6-fold increase in active caspase-6 and 12-fold increase in active caspase-3 levels were detected in N2a cells after treatment with staurosporine (Fig. 3B). We also set up a sensitive AlphaScreen assay to detect tau cleaved at Asp421. As shown in Fig. 3C, tau undergoes proteolytic cleavage at that site in a dose-dependent manner after treatment with staurosporine, and the maximum cleavage was reached at 0.25–1μM staurosporine after 24-h treatment. Tau cleavage at Asp421 above basal levels was confirmed by confocal microscopy (Fig. 3D). These results reassured us that the experimental set-up would allow us to look at the events of caspase activation and proteolytic processing of tau in the context of tau aggregation.

Treatment with staurosporine induces the MC1-conformational change of tau in N2a cells expressing human tau

Since we hypothesized that the induction of caspases required for apoptosis might affect the integrity and conformational status of tau, we measured the levels of total human tau and conformationally-changed tau in cell lysates collected after treatment of N2a cells with the well-known caspase activator staurosporine (Fig. 4). After treatment of N2a cells with staurosporine, the total levels of tau decreased simultaneously with increasing concentrations of the inducer, however at the same time the levels of conformationally-changed tau increased (Fig. 4B, C). As shown in Fig. 4C, there was a relative 1.6 to 26-fold increase of conformationally-changed tau after treatment with staurosporine.

Under our experimental conditions, the transient overexpression of tau and its conformational change did not seem to cause any detrimental effects on the nuclear compartment in non-staurosporine treated cells. However, enlarged images of cells with staining of total tau and MC1-tau after treatment with staurosporine indicate a change in morphology of the nucleus, and also tau localization around the nucleus (Fig. 4D).

Next, caspase activation and generation of tau conformational change were studied in a time-course experiment where N2a cells expressing human tau were treated with 1μM staurosporine (Fig. 5). Caspase activation was triggered within the first 6 h after the treatment and reached its maximal levels of 78% caspase positive cells after 24 h, with no significant induction of caspases in the non-treated cells (Fig. 5A, D). Also, there was a correlation between the induction of caspases by staurosporine treatment and the generation of conformationally-changed tau. As the levels of total tau decreased with the time of treatment, the levels of conformationally-changed tau increased, and after 24 h they were 29-fold higher than at the start of the experiment (Fig. 5E, F). During the 24-h incubation there was no change in total tau levels, but a 2.5-fold increase in conformationally-changed tau levels in the untreated controls (Fig. 5B, C). During the time course of staurosporine treatment there was an increase in membrane permeability of N2a cells treated with 1μM staurosporine for 6 h compared to untreated cells, as measured by LDH release. However, beyond that time point there was no difference in apparent toxicity between N2a cells treated with staurosporine versus the controls (Supplementary Figure 1B). This indicates that even though cell viability might be compromised to some extent upon increased caspase induction the severity of it does not worsen under our experimental conditions.

Both in dose-response and time-course experiments the levels of tau phosphorylation on Ser396/Ser404, Ser235, and Ser202 were decreased after treatment with staurosporine (Supplementary Figure 2B), as would be expected after treatment with a broad-spectrum kinase inhibitor.

The above results show that caspase activation can be induced by nanomolar concentrations of staurosporine and that there is a correlation between caspase activation triggered by staurosporine and generation of conformationally-changed tau, as measured in the MC1 ELISA assay.

Aggregates of tau are responsible for the enhanced signal observed in the MC1/CP27 ELISA format

To determine whether the tau conformational change triggered by staurosporine is accompanied by tau aggregation, we performed tau fractionation experiments. Cell lysates from N2a cells treated with staurosporine were subjected to ultracentrifugation in the presence of a non-ionic detergent and separated into soluble and insoluble fractions (Fig. 6A). There was a dose-dependent decrease in the levels of tau in the total and soluble fractions with a concomitant increase of insoluble tau. Staurosporine treatment also induced a pronounced pattern of tau cleavage products at all concentrations tested. These results are consistent with the results of the total tau analyses in the ELISA assay (Fig. 4A), where a dose-dependent decrease of total tau after treatment with staurosporine was seen. The accumulation of detergent-insoluble tau species in the insoluble fractions after treatment with staurosporine observed here follows a dose-dependent pattern, where the highest aggregation was observed with 0.5–1μM staurosporine. Using a total tau ELISA assay, tau was shown to be present in the low-speed, soluble and insoluble fractions. However, an ELISA assay testing for conformationally-changed tau only yielded strong positive signals for the low-speed and insoluble fractions while the soluble fraction gave a low range of absorbance signals above the background (Fig. 6B). This suggests that aggregates of tau are responsible for the enhanced signal observed in the MC1 ELISA assay. These aggregates were not stained by Thioflavin S, as assessed in confocal microscopy (data not shown), which imply that even though they are recognized by the conformational MC1 antibody they are likely not of a fibrillar nature. Follow-up analyses of aggresome formation revealed that the expression of tau itself induced aggresome formation by 8% above the basal levels, whereas treatment of N2a cells for 24 h with the proteasome inhibitor MG132 increased aggresome formation by 35% above the control (Supplementary Figure 3). Importantly, N2a cells expressing tau showed the presence of pronounced inclusions that co-localized with tau particularly after treatment with staurosporine already within 12 h of the experiment (Fig. 6C). These results indicate that tau species generated under conditions of staurosporine treatment become substrates for aggresome formation, and likely, clearance by lysosomes.

Caspase-driven proteolytic processing of tau and its aggregation occur within hours and can be blocked by caspase inhibitors

Having demonstrated the time- and concentration-dependent conformational changes of tau upon treatment with staurosporine and induction of caspases, we attempted to determine comparatively the timing of occurrence of tau proteolytic cleavage and its conformational change. Thus, we set up time-course and staurosporine dose-response experiments to follow in parallel tau cleavage at Asp421 and conformational change recognized by the MC1-antibody. Confirming the previous observations (Figs. 3 and 4), there was a dose-dependent increase in tau cleavage at Asp421 and generation of MC1-tau after treatment with staurosporine, but the pattern for both modifications differed over the time course of treatment (Fig. 7A, B). Interestingly, for all treatment conditions, the highest levels of Asp421-tau were observed as early as the 6-h time point, while prolonged incubation over 12 to 24 h led to loss of the epitope— most likely due to further proteolytic processing of the C-terminus (Fig. 7A). MC1-tau generation was triggered by 0.25μM and higher concentrations of staurosporine, and similar to the appearance of Asp421-cleaved tau it was already seen at the 6-h time point (Fig. 7B). The highest levels of MC1-tau were observed after 24-h treatment with 1μM staurosporine, at which point the levels of tau truncated at Asp421 were at their lowest under those experimental conditions. This indicates that the cleavage of tau at Asp421 is important in the cascade of tau conformational change, but after incorporation into aggregates the C-terminus of tau undergoes further proteolytic processing. To further test this hypothesis we assessed tau integrity using the pan-caspase inhibitor Z-VAD(Ome)-FMK in N2a cells transiently expressing tau and treated with increasing concentrations of staurosporine over 24 h. As expected, treatment of cells with staurosporine resulted in a dose- and time-dependent proteolytic cleavage of tau (Fig. 7C). After only 6 h of treatment, a pronounced pattern of tau degradation with multiple faster migrating fragments was observed, and after 24-h treatment with 0.25μM and 0.5μM staurosporine most tau was fully degraded. Interestingly, co-treatment of cells with staurosporine and pan-caspase inhibitor nearly completely halted tau degradation with treatment for 6 h in the range of staurosporine concentrations tested (Fig. 7C). The effect was even more dramatic at the 24-h time-point at which caspase inhibition significantly diminished tau proteolysis triggered by staurosporine.

We then looked at the levels of total and Asp421-caspase cleaved tau in N2a cells expressing human tau, treated with staurosporine and co-treated with the pan-caspase inhibitor. Levels of total tau and tau caspase-cleaved at Asp421 were measured in AlphaScreen assays (Fig. 7D, E). Here, the treatment of N2a cells with 0.158μM staurosporine did not cause a change in the total tau levels but 0.5μM staurosporine drove a dramatic 75% reduction of total tau. Levels of tau after treatment with pan-caspase inhibitor were restored to a higher degree than in the untreated cells, which implies that expression of tau itself may cause induction of caspases and tau degradation. Importantly, as shown in the Asp421-caspase cleaved tau AlphaScreen assay, tau cleavage was fully blocked by the pan-caspase inhibitor after treatment with both concentrations of staurosporine (Fig. 7E).

Next we tested whether tau conformational change and aggregation under our experimental conditions can be halted by treatment with the pan-caspase inhibitor ApoBlock. N2a cells transiently transfected with tau were treated with staurosporine to trigger caspase activation and consequent proteolytic cleavage of tau, its conformational change and aggregation (Fig. 8). As observed previously, the level of total tau is decreased 3-fold after treatment of N2a cells with staurosporine compared to the vehicle-treated cells. Strikingly, after pan-caspase inhibitor treatment the level of total tau increased by 64% compared to the vehicle-treated samples (Fig. 8A). There was a near 8-fold increase in conformationally changed tau (MC1-tau) with staurosporine treatment, which is in agreement with our previous observations. Crucially, the pan-caspase inhibitor reduced the levels of MC1-tau to 50% of the basal levels found in the control samples that were not treated with staurosporine (Fig. 8B). These data indicate that in our experimental system proteolytic processing of tau by caspases drives the observed conformational change, and inhibition of the induced and background levels of caspases can protect tau from proteolytic cleavage and conformational change as recognized by the MC1 antibody.