Common mouse models of tauopathy reflect early but not late human disease

Animals

P301S mice (Thy1-hTau.P301S (CBA.C57BL/6)) that were used for this study were homozygous for the MAPT*P301S transgene. The original breeder P301S mice were obtained from Medical Research Council (MRC) under licensing agreement. P301L mice [Tg(CamK2a-tTA)1Mmay Fgf14/Tg (tet0 MAPT*P301L) 4510Kha] were bigenic carrying a single copy of the CaMKIIα-tTA transgene and the MAPT*P301L transgene.

The P301L mice were generated by crossing animals hemizygous for the CaMKIIα-tTA transgene purchased from Jackson Laboratories (Stock Number 007004) to animals hemizygous for the MAPT*P301L transgene purchased from Jackson Laboratories under licensing agreement with Mayo Clinic (Stock Number 015815). Breeders for both lines were sent to Charles River Laboratories, Wilmington, MA for establishing and maintenance of mouse colonies as well as genotyping. Adult mice were then delivered to the AbbVie Cambridge Research Center (CRC) for use in studies. Once in the CRC vivarium, P301S and P301L mice were group-housed in individually ventilated cages (Innovive, San Diego, CA, USA) on a 12 h:12 h light–dark cycle and provided ad libitum access to food and water. All procedures conducted on animals at the CRC were approved by the AbbVie Institutional Animal Care and Use Committee (IACUC). To eliminate sex differences, only female mice were used for the study.

Tissue collection

To cover all stages of disease 5–7 biological replicates of 4–5 time points distributed over the whole disease progression were analyzed. This includes 2, 3, 4, and 5 months of age for the P301S and 1.5, 2.5, 4, 6, and 8 months of age for the P301L model. Mice were euthanized with sodium pentobarbital and perfused with 0.1 M phosphate-buffered saline. The brains were rapidly removed and right hemisphere per animal was dissected into selected brain regions. The brain regions collected, cortex and brainstem in P301S and cortex and hippocampus in P301L are known to show age-dependent increases in Tau pathology. For both models, the remaining unaffected regions, except for the cerebellum, were pooled and are referred to as the subcortical region. Tissues were placed in microfuge tubes and quickly frozen in liquid nitrogen. The left hemisphere was drop-fixed in 10% formalin for 24 h before being switched to 70% ethanol. Brains were stored in 70% ethanol until all brains from the study were collected to allow all samples to be processed for paraffin embedding at the same time (Leica).

Human tissue samples

Frozen human post-mortem frontal gyrus (BA46) specimens from patients carrying the P301L Tau mutation and healthy non-demented age-matched control subjects were obtained from the Massachusetts Alzheimer’s Disease Research Center Brain Bank. The demographic characteristics of the subjects are shown in Table S1.

Generation of heavy labeled 2N4R Tau standard

To quantify Tau amounts the FLEXITau workflow as previously described [21] was used. The heavy labeled Tau standard was transcribed and translated in vitro in a cell-free wheat germ expression system according to the manufacturer’s protocols (Cell Free Sciences, Wheat Germ Expression H Kit-NA) in the presence of heavy isotope 13C6-labeled lysine (+ 6) and 13C6,15N4-labeled arginine (+ 10). Afterward, the expressed Tau standard was dephosphorylated using Lambda Protein Phosphatase (New England Biolabs) according to the manufacturer’s instructions. Subsequent purification was performed using Ni-Sepharose beads (Ni-Sepharose High-Performance resin, GE Healthcare).

Fractionation of human and murine brain tissue samples

Frozen tissue samples (9–111 mg) were homogenized in 5 volumes TBS buffer (50 mM Tris–HCl buffer, pH 7.4, containing 150 mM NaCl, 0.5 mM MgSO4, phosphatase inhibitor cocktail (Roche), protease inhibitor cocktail (Roche) and Trichlostatin A (2µM)), using a Precellys tissue homogenizer (5500 rpm). To separate the insoluble Tau species from soluble Tau sarkosyl fractionation was performed on the homogenate. Therefore, cell debris was removed by centrifugation at 14,000 rpm for 20 min at 4 °C. The supernatant was diluted 1:1 with 2 × salt/sucrose solution (1.6 M NaCl, 20% Sucrose, 20 mM Tris–HCl buffer, pH 7.4, 2 mM EGTA, phosphatase inhibitor cocktail (Roche), protease inhibitor cocktail (Roche) and Trichlostatin A (2µM)). This supernatant is referred to as soluble fraction 1. Soluble fraction 2 was derived by reextraction of the remaining pellet with 1 × salt/sucrose solution (0.8 M NaCl, 10% Sucrose, 10 mM Tris–HCl buffer, pH 7.4, 1 mM EGTA phosphatase inhibitor cocktail (Roche), protease inhibitor cocktail (Roche) and Trichlostatin A (2µM)) and centrifugation at 14,000 rpm for 20 min at 4 °C. Both soluble fractions were treated with sarkosyl (1% final concentration) for 1.5 h at room temperature. Afterwards soluble fractions 1 and 2 were pooled and ultracentrifuged at 50,000 rpm for 1.5 h at 4 °C. The supernatant was transferred to a new tube (sarkosyl-soluble fraction). The sarkosyl-insoluble pellet, which contains the aggregated Tau species, was carefully washed twice with 20 µL PBS and resuspended in PBS using sonification (QSonica, 20% amplitude, 3 × 20 s). The protein concentration in the extracts was determined by bicinchoninic acid assay (BCA Protein Assay Kit, Thermo Scientific). After adding equal amounts of dephosphorylated and purified heavy 2N4R Tau standard the insoluble fractions were diluted with 8 M urea and processed separately using Filter Aided Sample Preparation (FASP Protein Digestion Kit, Expedeon) with DTT as reduction agent and 1% acrylamide for cysteine alkylation. Protein mixtures were digested with trypsin overnight at 37 °C (sequencing grade modified trypsin, Promega, Madison, WI). Acidified peptides were desalted using C18 extraction tips (Nest). Vacuum-dried peptides were reconstituted in sample buffer (0.1% formic acid, 5% acetonitrile).

FLEXITau measurement and analysis of human and murine sarkosyl fractions

Before Liquid chromatography-Selective Reaction Monitoring (LC-SRM) analysis, part of the resuspended sample was spiked with additional peptides. For the mouse models these included the light FLEX-peptide SENLYFQGDISR (15 fmol/µl final concentration), the heavy 0N Tau specific peptide STPTAEAEEAGIGDTPSL[+ 7]EDEAA[+ 4]GHVTQA[+ 4]R (50 fmol/µl final concentration) as well as the peptides carrying respective mutation (P301L: HVLGGGSVQIVYKPVDLSK[+ 8] or P301S: HVSGGGSVQIVYKPVDLSK[+ 8]). Human samples were spiked with heavy 0N, 1N (STPTAEAEEAGIGDTPSL[+ 7]EDEAA[+ 4]GHVTQA[+ 4]R), 3R (VQIVYKPVDLSK[+ 8]) and P301L Tau specific peptides (all spiked samples had a 50 fmol/µl final concentration) (all peptides were in QuantPro quality synthesized by Thermo Fischer). LC-SRM measurements of Tau L/H peptide ratios were performed as described previously [21]. After optimization of transitions using in-house DDA spectral libraries and heavy-isotope labeled Tau standards the samples were analyzed on a quadrupole mass spectrometer (5500 QTRAP, Sciex) which was coupled to an Eksigent micro-autosampler AS2 and a microflow pump (Eksigent/Sciex, Framingham, USA) as described above but operated at 5 µL/min. Here, 1.25 µg of peptides seperated on a 25 cm column (Proteocol C18G 200A˚, 250 mm × 300 μm ID Trajan Scientific and Medical, Australia) using a 25 min gradient from 0 to 35% acetonitrile. Three to five transitions were monitored for each precursor by SRM with a retention time window of 45 s and a target scan time of 0.5 s to ensure an optimal amount of data points per peak. SRM data were analyzed and validated in Skyline-daily (version 21.0.9.105, MacCoss Lab Software, University of Washington, Seattle, WA) [22]. In total 18 Tau peptides were quantified, and the L/H ratios of the peak area were exported for every peptide. The absolute abundance of Tau was calculated using the FLEX peptide L/H ratio and the L/H ratio of the peptide with the highest ratio that is shared between the mouse and the transgene human Tau as described before [23]. The modifications extent was calculated through normalization of the peptide L/H ratio to the shared or human-specific peptide with the highest L/H ratio, depending on the specificity of the peptide. Plotting, student’s t-test and Pearson correlation were done using GraphPad Prism 8 version 8.2.1 (GraphPad Software Inc.). Hierarchical cluster analysis using euclidean distance and complete linkage was performed in R (4.1.2) using R studio (2021.09.1) and the pheatmap (1.0.12) package [24, 25].

LC–MS/MS of murine sarkosyl-insoluble fractions and data analysis

The remaining part of the sample buffer resuspended murine sample was used for LC–MS/MS analysis. Here a QExactive mass spectrometer (Thermo Fisher Scientific, Bremen) coupled to a micro-autosampler AS2 and a nanoflow HPLC pump (Eksigent, Dublin, CA) was used. Peptides were loaded on a capflow PicoChip column (150 mm × 10 cm Acquity BEH C18 1.7 mm 130 Å, New Objective, Woburn, MA) with 2 ml/min solvent A (water + 0.1% formic acid). The elution was performed by a 135 min gradient at a flow rate of 1 µl/min. Solvent B (acetonitrile + 0.1% formic acid) was increased from 2 to 20% over the first 110 min. Between 110 and 120 min, it was further increased to 30%. For the wash step solvent B was ramped up to 95% within 1 min and was kept constant at this percentage for 5 min. Afterwards, a re-eqilibration step at 2% B for 5 min was performed. During the whole run, the PicoChip containing an emitter for nanospray ionization was kept at 50 °C. A full mass spectrum with a resolution of 70,000 was acquired in a mass range of 375–1400 m/z (AGC target 3 × 106, maximum injection time 60 ms). The 12 most intense precursor ions were selected for fragmentation via higher-energy c-trap dissociation (HCD, resolution 17,500, AGC target 5×104, maximum injection time 100 ms, isolation window 1.6 m/z, normalized collision energy 27%). Once a precursor ion was picked for fragmentation it was excluded for the following 25 s.

To identify PTMs of Tau, the MS raw data were processed with ProteinPilot™ Software 5.02 (Paragon Algorithm 5.0.2.0.5174, Sciex), MaxQuant software version 1.6.5.0 [26], and Mascot using the Mascot Deamon version 2.6.0 (Matrix Science). QExactive raw files were converted into mgf data. Collected spectra were searched against a mus musculus proteome database including isoforms (21,215 entries, downloaded from uniprot.org on 06/10/2019) which was used in all three search engines. To avoid mismatching of PTMs all murine Tau isoforms were removed and the human 0N4R Tau (Uniprot ID: P10636-6) with the P301S (dbSNP ID: rs63751438) or P301L (dbSNP ID: rs63751273) mutation was added.

In ProteinPilot™ the following settings were applied: sample type ‘Identification’; Cys Alkylation ‘Propionamide’; Digestion ‘Trypsin’; instrument type ‘Orbi MS, Orbi MS/MS’; ‘thorough ID’ search mode; ‘ID focus on biological modifications’. A cutoff of 95% confidence was employed for all modified peptides. In addition, only phosphorylation of S, T and Y, methylation of K and R, acetylation of K and ubiquitination of K were considered.

Within the MaxQuant software, the following settings were used: trypsin (specificity set as Trypsin/P) with up to two missed cleavages and a minimum peptide length of 5 amino acids. Oxidation of M, acetylation of N-termini and K, phosphorylation of S and T, methylation of K and R and ubiquitination (GlyGly) of K were chosen as variable modifications and propionamide was set as static modification of cysteine with a maximum of three modifications per peptide. False discovery rate (FDR) was set to 1% on peptide and protein levels and was determined by searching a reverse database. Peptide identification by match between runs was disabled. For all other search parameters, the default settings were used.

The Mascot search was performed considering peptide charge states of 2 + , 3 + and 4 + including a 10 ppm tolerance. The MS/MS search was run with a mass tolerance of 0.6 Da. The search was performed with trypsin as the used enzyme allowing a maximum of 2 missed cleavages and Oxidation of M, acetylation of K, methylation of K and R, citrullination of R, phosphorylation of S, T and Y and ubiquitination (GlyGly) of K were chosen as variable modifications and propionamide was set as static modification of cysteine. Afterwards, the PTM results of different search algorithms were cumulatively combined. Hierarchical cluster analysis using euclidean distance and complete linkage was performed in R (4.1.2) using R studio (2021.09.1) and the pheatmap (1.0.12) package [24, 25].

LC–MS/MS of human sarkosyl-insoluble fractions and data analysis

The remaining part of the sample buffer resuspended human sample was used for LC–MS/MS analysis. Here a timsTOF Pro mass spectrometer (Bruker Daltonics, Billerica, MA) coupled to a nano elute liquid chromatography (Bruker) was used. Peptides were loaded on a C18 UHPLC column 25 cm × 75 μm (1.6 µm particle size) from IonOpticks (Fitzroy, Australia). The elution was performed by a 120 min gradient at a flow rate of 0.4 µl/min. Solvent B (acetonitrile + 0.1% formic acid) was increased from 0 to 23% over the first 90 min. Between 90 and 100 min, it was further increased to 35%. For the wash step solvent B was ramped up to 80% within 10 min and was kept constant at this percentage for 10 min. During the whole run, the column was kept at 50 °C. For the data-dependent analysis, the mass spectrometer was operated in DDA-PASEF mode. 10 PASEF MS/MS scans were triggered per cycle. DDA-PASEF parameters were set as follow: m/z range 100–1700, mobility (1/K0) range was set to 0.60–1.6 V.s/cm2, the accumulation and ramp time were of 100 ms. Target intensity per individual PASEF precursor was set to 20,000. The values for mobility-dependent collision energy ramping were set to 59 eV at an inversed reduced mobility (1/K0) of 1.6 V.s/cm2 and 20 eV at 0.6 V.s/cm2. Collision energies were linearly interpolated between these two 1/K0 values. The acquired data was converted to mgf using the Compass data analysis software (Bruker, version 5.3).

For identification spectra were searched against a canonical homo sapiens proteome database (20,370 entries, downloaded from uniprot.org on 11/24/2020) which was used in all search engines. To avoid mismatching of PTMs all 6 human Tau isoforms and all 4R Tau isoforms including the P301L (dbSNP ID: rs63751273) mutation were added to the database.

Identification of peptides was performed using the Mascot and Fragpipe search algorithm. The Mascot search was performed the same way as the murine samples. The Fragpipe search included the MSFragger, Philosopher and IonQuant modules [27,28,29,30]. MSFragger 3.4 was ran using the standard settings. Oxidation of M, acetylation of K, methylation of K and R, phosphorylation of S, T and Y and ubiquitination (GlyGly) of K were chosen as variable modifications and propionamide was set as static modification of cysteine. Philosopher 4.1.1 was used for statistical validation of identified peptides. IonQuant 1.7.17 was used for quantification where a minimum of 1 ion was used for peptide quantification. Afterwards, the PTM results from different search algorithms were cumulatively combined. Hierarchical cluster analysis using euclidean distance and complete linkage was performed in R (4.1.2) using R studio (2021.09.1) and the pheatmap (1.0.12) package [24, 25].

Tissue processing and AT100 immunoreactivity

Three to four brains from different treatment groups were embedded together into paraffin blocks (Sakura Finetek, USA) in sagittal plane. Five µm paraffin sections were collected through the entire brain and mounted onto a glass slides. Matched sections through the cortex and brainstem for P301S mice and cortex and hippocampus for the P301L mice were stained for AT100 immunoreactivity using the BOND RX stainer with the Refine Detection kit (Leica). In brief, the slides were deparaffinized, rehydrated, and placed on the BOND RX where they underwent a series of pretreatments including peroxide block (Leica) 5% donkey serum blocking (017–000-121, Jackson ImmunoResearch Labs), mouse IgG blocking (MKB-2213, Vector laboratories) and were then incubated overnight at 4 C° in the mouse monoclonal antibody to AT100 tau (MN1060, Thermo Fisher) at a concentration of 0.006 µg/mL. Slides were then returned to the BOND RX where they underwent series of washes before being incubated in Biotin-SP-conjugated F(ab')2 donkey anti mouse IgG(H + L) secondary antibody (715–006-151, Jackson ImmunoResearch Labs, West Grove, PA, USA) at a concentration of 2 µg/mL for 20 min. After additional washing steps, the slides were incubated for 15 min in Streptavidin/Horseradish Peroxidase (RE7104, Leica) and the immunoreactivity was visualized using diaminobenzidine (DAB; Leica, Buffalo Grove, IL, USA) and counterstained with hemotoxylin (Leica).

AT100 immunoreactivity quantification

For both studies, 3 matched sections per brain were analyzed using the Area Quantification module in HALOTM image analysis software (Indica Labs, Albuquerque, NM, USA). For P301s and P301L mice, the entire cortex was outlined as the region of interest. Using the software, the "threshold" was determined by an observer who was blind to the age of the animals. The threshold was set so that the positive brown DAB stain of the AT100 immunoreactivity was recognized by the software and the background/non-specific staining was excluded from the analysis. Once an appropriate threshold was set, the software measured the percentage of the area of interest (cortex) containing the positive immunoreactivity (AT100). The percentages for the 3 sections measured per brain were then averaged to obtain a single value for each animal.

留言 (0)

沒有登入
gif