Tau as a mediator of neurotoxicity associated to cerebral amyloid angiopathy

ADan peptide (EASNCFAIRHFENKFAVETLICFNLFLNSQEKHY) [63] was synthesized by ThermoFisher Scientific using Fmoc-based Solid Phase Peptide Synthesis and purified by HPLC. To prepare ADan oligomers, ADan peptide was resuspended in PBS without calcium and magnesium to a final concentration of 0.5 mg/mL. The ADan solution was then stirred at room temperature (RT) for 48 h. Aliquots were collected at different time points and stored at − 80 °C. To confirm the formation of ADan oligomers, Western Blot (WB) analysis was performed using anti-ADan 1699 antibody (1:1000, developed by R. Vidal) that was specific for residues 23–34 (FNLFLNSQEKHY) of the ADan amyloid peptide [59] and the conformational antibody anti-oligomers F11G3 (1:1000, provided by R. Kayed), as previously described [44].

Human embryonic kidney cells expressing doxycycline (Dox) inducible full length human tau (2N4R) with the P301L mutation (HEK P301L) ([16] and Additional file 1: Figure S1) were cultured in DMEM (Invitrogen) with 10% fetal bovine serum (Invitrogen), 2 mM glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin. 1 μg/mL Dox was added into the culture media to induce human tau expression 24 h prior to transfection and ADan oligomer treatment. Human wild-type (WT) BRI₂ and BRI₂ bearing the Danish mutation were cloned into a pcDNA3.1 vector (Invitrogen). The sequences were confirmed by Sanger sequencing. All plasmids were transfected with Lipofectamine 2000 (Invitrogen) and incubated for 48 h. For ADan oligomer treatment, human Tau expression was induced by 1 μg/mL Dox in HEK P301L cells for 24 h, the cells were treated with 200 nM ADan oligomers, monomers, or PBS for 8 h. All measurements were made in triplicates.

HEK P301L cells were plated at 20,000 cells per well in 96-well plates with 1 μg/mL Dox for 24 h. Then the cells were treated with 2.5 μM ADan oligomers, monomers, or PBS. After incubation for 4 h at 37 °C, cell viability was assayed using a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) cell proliferation assay kit (Promega) according to the manufacturer’s specifications. The MTT assay is a colorimetric assay for assessing cell metabolic activity, which reflects the number of viable cells. All measurements were made in six replicates.

Cells were washed twice with 0.5 mL PBS and lysed on ice for 60 min in 80 μL T-PER tissue protein extraction reagent (ThermoFisher Scientific) supplemented with Complete protease inhibitors cocktail (Roche). The cell lysates were then centrifuged at 14,800 rpm for 15 min at 4 °C, the supernatants were analyzed by WB as previously described [44]. Primary antibodies used were anti-tau (tau-5, 1:1500, Abcam), anti-ptau Thr231 (1:1500, Millipore), PHF1 anti-ptau Ser396/Ser404 (1:1000, provided by Dr. Peter Davies), Anti-BRI₂ (1:100, Santa Cruz Biotech) and Anti-GAPDH (1:10000, Sigma-Aldrich). Tau oligomers levels were measured using the anti-tau oligomer antibody T22 (1:1000, provided by Dr. Rakez Kayed) as previously described [43].

HEK P301L cells were plated on the 12 mm Poly-L-Lysine coated coverslips (Corning BioCoat) in 24-well plates at a density of 62,500 cells per well with 1 μg/mL Dox for 24 h, then cells were transfected with BRI₂ bearing the Danish mutation for 48 h. Then, the coverslips were briefly rinsed with warm TBS and fixed with 2% paraformaldehyde for 10 min at RT, then permeabilized for 5 min with 0.1% Triton X-100. The samples were blocked for 1 h in 5% goat serum and incubated overnight (O.N.) at 4 °C with mouse anti-ptau Thr231 (1:1000, Millipore) and rabbit anti-ADan antibody 1699 (1:500). Cells were then washed in TBS, then incubated with Alexa 488-conjugated goat anti-mouse antibody (1:700, Invitrogen) and Alexa 568-conjugated goat anti-rabbit antibody (1:700, Invitrogen) for 1 h at RT, then washed with TBS and mounted with Vectashield mounting medium with DAPI (Vector Laboratories). Samples were examined using a Leica DMi 8 epifluorescence microscope coupled with the LAS X program (Leica). For orthogonal images of reconstructed three-dimensional views a Nikon A1-R Laser Scanning Confocal Microscope coupled with the NIS Element Advanced Research software was utilized.

Tg-FDD, wild type C57/BL6J (WT) (JAX stock #000664) and Tau knock out (Tau^−/−) (JAX stock #007251) male and female mice were used for our experiments, including cellular, biochemical, and immunohistochemistry (IHC) analyses. The Tg-FDD mouse model [59] expresses a FDD-associated human mutant BRI₂ transgene. Mice were housed at the Indiana University School of Medicine (IUSM) animal care facility and were maintained according to USDA standards (12-h light/dark cycle, food and water ad libitum), in accordance with the Guide for the Care and Use of Laboratory Animals (National Institutes of Health, Bethesda, MD). Animals were anesthetized and euthanized according to IUSM Institutional Animal Care and Use Committee-approved procedures. Tissue was collected from 3 and 18 months old animals (5 mice per genotype). Mice were deeply anesthetized and perfused transcardially with PBS prior to decapitation. After sacrifice, brains were removed and stored at − 80 °C or formalin fixed as previously described [59].

Tg-FDD and WT (5 mice per genotype) brains were homogenized at a 1:10 (w/vol) ratio of brain and T-PER tissue protein extraction reagent with complete protease inhibitor cocktail (Roche). Samples were then centrifuged at 13,200 rpm for 15 min at 4 °C. The supernatants were aliquoted, snap-frozen, and stored at − 80 °C until analyzed. The T-PER insoluble pellets were resuspended in 88% formic acid (FA) at one fourth volume of their brain homogenates, then incubated for 1 h at RT. Samples were then diluted with distilled water to obtain the same volume used for brain homogenates. Samples were then lyophilized for 24 h. Freeze-dried samples were reconstituted in PBS using the same volume as brain homogenates, then sonicated for 30 s. Finally, samples were mixed with sample loading buffer, run on a NuPAGE 4–12% Bis-Tris protein gel (Invitrogen), and analyzed by WB. Primary antibodies used were anti-BRI₂ (1:100, Santa Cruz Biotech), Tau-5 (1:1500, Abcam), PHF1 (anti-ptau Ser396/Ser404) (1:1000, gift from Peter Davies), anti-ptau Thr231 (1:1500, Abcam), MC1 (1:100, gift from Peter Davies), and anti-Vinculin (1:10000, Invitrogen). Secondary antibodies used: goat anti-mouse HRP IgG (1:1500, Invitrogen) and goat anti-rabbit HRP IgG (1:1500, Invitrogen). WB quantification results are expressed as the ratio of phospho-tau to total tau levels. Then this value was normalized by the loading control. In all cases, we considered the control group as the 100%. For the FA treated samples, equal amount of proteins were loaded, previously determined by Bradford assay (ThermoFisher Scientific) according to manufacturer’s specification.

Paraffin sections were deparaffinized in xylene and rehydrated in ethanol (EtOH) and washed with deionized water. Then the sections were heated with a microwave oven in low pH antigen retrieval solution (eBioscience) twice for 4 min each. After washing in TBS twice for 5 min each, the sections were blocked in 5% normal goat serum for 1 h at RT, sections were incubated O.N. at 4 °C with the following antibodies: anti-ADan 1699 (1:500), T22 (1:100, gift from Dr. Rakez Kayed), TOMA (1:750, gift from Dr. Rakez Kayed), MC1 (1:750, gift from Dr. Peter Davies), anti-GFAP antibody (1:200, Sigma-Aldrich), anti-Iba1 (1:250, Millipore), and NeuN (1:100, Abcam). For amyloid detection, sections were first stained with 1% Thioflavin S (Thio-S) in TBS for 8 min before the primary antibody incubation, then washed in EtOH and deionized water. The next day, the sections were washed in TBS and incubated with secondary antibodies Alexa 488-conjugated goat anti-mouse antibody (1:500, Invitrogen) and Alexa 568-conjugated goat anti-rabbit antibody (1:500, Invitrogen) for 1 h at RT. Finally, sections were washed in TBS and mounted with Vectashield mounting medium with DAPI (Vector Laboratories). For colocalization analysis we utilized the Coloc 2 plugin from Fiji that implements and performs the pixel intensity correlation over space methods of Pearson [55]. For every single staining, as a negative control, primary antibodies were omitted to determine background and autofluorescence (not shown). WT and Tg-FDD cerebral cortex was examined using a Leica DMi 8 epifluorescence microscope coupled with the LAS X program (Leica).

Mouse brains total RNA were isolated with RNeasy Plus Universal Mini kit (Qiagen). cDNA was prepared from 1 μg total RNA with High-Capacity cDNA reverse transcription kit (Life Technologies). All qPCRs were performed on QuantStudio 6 Flex Real-Time PCR system (Life Technologies). The mouse Mapt relative gene expression was evaluated with delta Ct method using Taqman probe sets (Mapt: Mm00521988_m1, GAPDH: 4351309, Life Technologies) and TaqMan Universal PCR Master Mix.

Immediately following euthanasia via decapitation under deep isoflurane anesthesia, the brain was quickly excised and placed in an ice-cold tissue cutting solution containing (in mM): 194 sucrose, 30 NaCl, 4.5 KCl, 1 MgCl₂, 26 NaHCO₃, 1.2 NaH₂PO₄, 10 Glucose saturated with a mixture of 95% O2 and 5% CO2, and sliced to a thickness of 280 μm on a vibratome (Leica VT1200S, Germany). Slices were transferred to an artificial cerebrospinal fluid (aCSF) solution containing (in mM): 124 NaCl, 4.5 KCl, 1 MgCl₂, 26 NaHCO₃, 1.2 NaH₂PO₄, 10 Glucose, 2 CaCl₂ (310–320 mOsm) saturated with 95% O2/5% CO2 at 30 °C for 1 h before being moved to RT. When ready for recording, in control conditions, slices were transferred to a recording chamber continuously perfused with aCSF solution saturated with 95% O2/5% CO2. In experiments using the ADan oligomer, the slices were first transferred to an incubation chamber containing aCSF containing 400 nM ADan oligomer saturated with 95% O2/5% CO2 at RT for at least 1 h before being transferred to the recording chamber.

Cerebellums were obtained from 8 to 10 days old mouse pups from WT C57/BL6J (JAX stock #000664) and Tau knock out (JAX stock #007251) mice. Their brains were dissected, the cerebellum separated, cleaned thoroughly from meninges, and incubated with 0.25% trypsin EDTA and 25 mg/mL DNase. Then they were disaggregated by passing them through a 22^1/2 G syringe needle previously coated with 100 mg/mL BSA in ultrapure cell culture grade water. The suspension was centrifuged for 5 min at 1250 rpm, and then cells were resuspended in 8 mg/mL BSA. A second centrifugation was followed by a resuspension in 2 mg/mL BSA. Finally, cells were resuspended a third time in primary granule neuron media (1X Neurobasal medium, Invitrogen) and counted. They were seeded in 24-well plates that contained coverslips previously treated with 0.01% poly-L-lysine. Seeding number was 50,000 cells/well. After 14 days in vitro (DIV), cells were fixed for 15 min at 37 °C in a solution of 4% paraformaldehyde/sucrose in TBS. Cells were washed with TBS, permeabilized with 0.25% Triton X-100 in TBS, and blocked with 10% BSA in TBS. Primary antibodies anti-Synapsin 1 (1:100, Abcam) and anti-PSD95 (1:100, Abcam) were incubated O.N. at 4 °C. The next day cells were washed in TBS, permeabilized in 0.25% Triton X-100, and incubated with secondary antibodies Alexa Fluor-488 anti-mouse (1:100, Invitrogen) and Alexa Fluor-568 anti-rabbit (1:100, Invitrogen) for 1 h at RT. Finally, cells were washed in TBS for 3 min, mounted with Vectashield mounting medium with DAPI, and visualized using a Leica DMi 8 epifluorescence microscope coupled with the LAS X program (Leica). Clusters of Synapsin 1 and PSD95 were quantified as previously described [14].

Field excitatory postsynaptic potential (fEPSP) recordings from the stratum radiatum of hippocampal area CA1 were carried out at 29–32 °C and aCSF was continuously perfused at a rate of 1–2 ml/minute using a Multiclamp 700B amplifier (Axon Instruments, Union City, CA) signals were amplified (gain 100) and filtered (1 kHz), then digitized (10 kHz). Slices were visualized on an Olympus BX51WI microscope (Olympus Corporation of America). Picrotoxin (50 μM) was added to the aCSF for recordings to isolate excitatory potentials. Micropipettes were prepared from filament-containing borosilicate micropipettes (World Precision Instruments) using a P-1000 micropipette puller (Sutter Instruments, Novato, CA), having a 2.0–3.5 MΩ resistance. The glass micropipettes were filled with 1 M NaCl and placed into the stratum radiatum of hippocampal area CA1. A concentric bipolar stimulating electrode (FHC, Bowdoin, ME) was placed into stratum radiatum in area CA1 ∼ 500 μm from the recording site. fEPSPs were generated by a DS3 Isolated Current Stimulator (Digitimer, Ft. Lauderdale, FL) every 20 s and stimulus intensity was adjusted to produce stable fEPSP responses prior to the initiation of experimental recording. A 10 min baseline was recoded before delivery of four high frequency stimulations (HFS; 100 Hz, 1 s duration, 10 s inter-stimulus-interval). Data were acquired using Clampex 10.3 (Molecular Devices, Sunnyvale, CA). The representative traces were obtained from the average baseline fEPSP (1–10 min) and average post-HFS fEPSP of the final 10 min of recording. Exclusion of individual data points was determined using an outlier calculator included in the Prism 7 software package. Recordings were made 2–7 h after euthanasia. The analysis of fEPSP slopes were used to determine the effects of ADan oligomer. The experimenter was not blinded to treatments administered to the slices.

Experimental analysis and data collection were performed in a blinded fashion, if not stated otherwise. P values were determined using the appropriate statistical method via GraphPad Prism, as described throughout the manuscript. Statistical comparisons were made using two-tailed unpaired Student’s t-test with Welch’s correction, Mann-Whitney test or one-Way ANOVA with Tukey’s correction. Figures 1b, 2d, 3b-d, and 8C use unpaired Student’s t-test follow by Welch’s correction. Figure 7b-d use Mann-Whitney test and Fig. 2c use one-Way ANOVA followed by Tukey’s correction. Data are presented as mean ± SD if not stated otherwise. *, **, ***, and **** denote P < 0.05, P < 0.01, P < 0.001, and P < 0.0001 respectively. All details of experiments can be found in the Results or the Figure Legends.

FDD patients are not only characterized by the vascular accumulation of ADan amyloid, but also by the accumulation of hyperphosphorylated tau. Therefore, we decided to determine in a cellular model if ADan promotes tau hyperphosphorylation. To do so, we transfected BRI₂ bearing the Danish mutation in HEK cells, lacking endogenous tau, that conditionally express human tau with the P301L mutation after the addition of Dox [16]. As a control, we transfected WT BRI₂. Danish mutant BRI₂ overexpression led to an increase in the levels of phosphorylated tau at Ser396/Ser404 and Thr231 (Fig. 1a-b). To determine if the effect of ADan aggregates over tau phosphorylation and aggregation is mediated by a direct interaction between ADan peptides and tau or by an indirect mechanism, we performed double-immunostaining using an anti-ADan antibody and an anti-phospho tau antibody (Thr231). We observed colocalization of ADan immunoreactivity with phospho-tau Thr231 in a number of cases, but this was not observed in all cases (Fig. 1c). Overall, these results suggested that the expression of BRI₂ bearing the Danish mutation could induce the aggregation and phosphorylation of tau, a process that, in cell culture, could be mediated by a direct event or by an indirect mechanism.

Since BRI₂ is a transmembrane protein and the Danish mutation induced the secretion of the ADan peptide to the extracellular space [22] (Additional file 1: Figure S2), we evaluated the effect of ADan monomers and oligomers over tau phosphorylation and cellular toxicity. To test this, we prepared ADan recombinant oligomers as we previously described [44]. Figure 2a shows the formation of ADan oligomers immunoreactive with the conformational anti-oligomer F11G3 antibody after 48 h of stirring. Incubation of ADan oligomers or monomers with HEK cells that conditionally express mutant tau for 8 h showed that when cells are exposed to ADan oligomers, tau becomes hyperphosphorylated (Fig. 2b-c). These ADan oligomers and even monomers, in a lower degree, are able to get internalized into the cells (Additional file 1: Figure S3). Moreover, cells treated with oligomers for 4 h showed a significant cellular toxicity as measured by the MTT cell proliferation assay (Fig. 2d). Surprisingly, ADan oligomer toxicity was only observed when cells were induced to express human tau after the addition of Dox, but not when cells did not express it (Fig. 2d). Overall, these results suggest that ADan oligomers could induce tau hyperphosphorylation and subsequent tau-dependent toxicity.

We tested whether ADan aggregates affect endogenous murine tau phosphorylation in vivo in the Tg-FDD model, which is characterized by ADan amyloid deposits in the vasculature [59] (Additional file 1: Figure S4). WB analysis of soluble brain fractions showed a significant increase in phosphorylated tau at Ser396/Ser404 and Thr231 in 18 months old Tg-FDD mice compared to WT controls of the same ages (Fig. 3a-c). No changes in the levels of phosphorylated tau at Ser202/Thr205, Ser214, Ser262 and Ser356 were observed (data not shown). When we performed WB analysis using the MC1 antibody that recognizes early stages of tau misfolding, we observed a significant increase in MC1-positive tau (Fig. 3a and d), suggesting that the accumulation of vascular amyloid in the Tg-FDD model promotes murine tau misfolding. Quantification of murine tau mRNA levels did not show any differences between Tg-FDD and WT mice (Fig. 3e), demonstrating that the effect of ADan amyloid in the Tg-FDD model over tau is at the protein level. We then performed biochemical characterizations of the insoluble brain fractions from WT and Tg-FDD mice. WB analysis of these fractions demonstrated the presence of insoluble ADan in the Tg-FDD mice (Fig. 3f), but not tau (Fig. 3f) or phospho-tau (data not shown), suggesting that in the Tg-FDD model endogenous murine tau does not accumulate into insoluble aggregates. Interestingly, whenever we performed WB analysis of soluble brain fractions from 3 months old Tg-FDD mice, an age when no vascular amyloid is observed [59], no changes in the level of phosphorylated tau were observed (Additional file 1: Figure S5). Overall, these results suggest that vascular amyloid accumulation possibly induces endogenous tau phosphorylation and misfolding.

To confirm the amyloidogenic nature of ADan deposits in the vasculature of the Tg-FDD model, we double stained brain sections of 18 months old mice with the ADan antibody and Thio-S. We observed that ADan deposition in the cerebral vessels was mainly Thio-S-positive (Fig. 4a-c). Interestingly, small ADan depositions negative for Thio-S were also observed (Fig. 4a-c). To determine the presence of ADan soluble aggregates, such as oligomers, we performed double immunofluorescence in brain sections using the ADan antibody and the conformation specific monoclonal anti-oligomer F11G3 antibody (Fig. 4d-f) [31, 44]. Substantial intracellular co-localization between both antibody signals was observed (Fig. 4f).

We performed double immunofluorescence in adjacent sections using anti-ADan and MC1 or TOMA antibodies to assess the spatial relationship between vascular ADan deposits and tau. Both MC1 and TOMA antibodies recognized early stages of tau aggregation such as oligomers [11, 37]. Double staining of anti-ADan and TOMA on Tg-FDD sections showed the accumulation of tau oligomers in the vicinity of vessels enriched with amyloid deposits in the cortex (Fig. 5a, top and bottom panel). Double staining of anti-ADan and MC1 confirmed the presence of misfolded tau surrounding ADan-positive vessels (Fig. 5b, top panel). Interestingly, double staining of anti-ADan and MC1 also revealed intracellular association between ADan and MC1-positive tau in the cerebral cortex (Fig. 5b, bottom panel). Overall these results demonstrated the presence of two types of pathological accumulation of tau in this CAA model. One where misfolded tau accumulates in the vicinity of vessels enriched with ADan deposits, and a second one where misfolded tau co-localized with ADan intracellular signal in areas of the cerebral cortex, where no major vascular amyloid deposits are observed. Tau oligomers were observed in other brain regions (Additional file 1: Figure S6) that have been previous associated with the accumulation of vascular amyloid in this mouse model [59].

Innate immunity is activated both in cerebrovascular disease [38] and in AD. In the context of our CAA model, we observed severe astro and microgliosis, as evidenced by robust GFAP and Iba1 staining in areas affected by ADan amyloid vascular deposits (Additional file 1: Figure S7). Based on this observation, we decided to determine which cell type showed tau oligomers associated with vascular amyloid. We performed triple immunofluorescence in Tg-FDD brain sections for Thio-S (amyloid), T22 (tau oligomers), and GFAP or Iba1 or NeuN for astrocytes, microglia, or neurons, respectively. Fluorescent microscopy suggests that tau oligomers are closely associated with GFAP-positive astrocytes (Fig. 6a-e). Co-localization analysis between T22 and GFAP signal suggest that these oligomers are located in the cytoplasm and astrocytic processes (Fig. 6e). Triple staining with Thio-S, T22, and Iba1 showed that tau oligomers were not associated with Iba1-positive microglia (Fig. 6f-j). Interestingly, triple staining revealed that tau oligomers are associated with neurons in areas throughout the cortex where no Thio-S-positive vascular structures were located (Fig. 6k-o). Substantial co-localization between T22 and NeuN antibody signals was observed (Fig. 6o).

It is widely accepted that the adverse effects of Aβ on neuronal degeneration and cognitive dysfunction appear to depend largely on soluble tau [29]. For instance, it has been reported that the reduction of tau expression prevents Aβ-induced neurodegeneration in cell culture [51, 65]. Moreover, in vivo studies have shown how genetic ablation of endogenous murine tau in a hAβPP mouse model prevents behavioral deficits without altering the amount of Aβ plaques in the parenchyma [54]. Therefore, considering these previous studies, in addition to the fact that a decrease in the levels of synaptic markers in Tg-FDD mice has been reported [23], we aimed to investigate if the neuronal damage triggered by ADan, a highly vasculotropic amyloid, is also dependent on tau.

To investigate if ADan aggregates induce impairment of the synaptic structure via tau, we performed IF for the pre- and post-synaptic markers Synapsin-1 and PSD95 respectively, and assessed their localization at synapses in cultured primary cerebellar granule neurons from WT mice and mice with a genetic knock-out of Mapt gene (Tau^−/−) in the presence or absence of ADan oligomers (Fig. 7a). Our results show a decrease in the total number of clusters or puncta marked for Synapsin-1 and PSD95 in WT neurons incubated with 500 nM of ADan oligomers in comparison with untreated WT control groups. Remarkably, ADan oligomers did not have any effect on the number of clusters marked either for Synapsin-1 or PSD95 on Tau^−/− neurons (Fig. 7b-c). ADan oligomers not only affected the number of Synapsin-1 and PSD95 clusters in WT neurons, but also has a detrimental effect on the colocalization between these pre- and post-synaptic markers. On the other hand, no effect on the synaptic colocalization between Synapsin-1 and PSD-95 was observed in Tau^−/− neurons after the addition of ADan oligomers (Fig. 7d). To correlate these results with a functional analysis, we performed hippocampal long-term potentiation (LTP) experiments in hippocampal area CA1 in WT and Tau^−/− mouse-derived brain slice preparations. We monitored fEPSPs evoked by extracellular stimulation of the Schaffer collateral pathway and induced LTP using high-frequency stimulation (four stimuli of 100 Hz for 1 s with a 10 s inter-stimulus interval). In WT slices pretreated with 400 nM ADan oligomers for 2 h, LTP was almost completely blocked (Fig. 8a and c). This was a specific effect of ADan oligomers, because control slices showed LTP (Fig. 8a and c). These results demonstrate that, in WT mice, acute application of ADan impairs one or more of the cellular mechanisms necessary for LTP. To confirm a possible functional interaction between ADan and tau, we tested the effect of ADan oligomers on LTP in Tau^−/− mice. Slices incubated in the control solution showed normal levels of LTP and, remarkably, slices preincubated in ADan oligomers showed LTP levels of similar magnitude (Fig. 8b-c).

Overall, these results indicate that the absence of tau could prevent the synaptic dysfunction induced by a CAA-associated amyloid such as ADan.