Male C57BL/6 mice were purchased from Jackson Laboratories (stock number 000664) and used for ASO treatment studies at 4 months of age. hTau transgenic mice (Andorfer et al., 2003) and PS19 transgenic mice (Yoshiyama et al., 2007) were bred in-house and identified for transgene expression as previously described (Schoch et al., 2016; DeVos et al., 2017). For ASO treatment studies, we used male and female hTau mice at approximately 9 months of age and male and female PS19 mice and non-transgenic littermates at approximately 6 months of age. Animals were maintained on a 12-h light/dark cycle and provided ad libitum access to food and water. All husbandry and experimental procedures were approved by the Institutional Animal Care and Use Committee at Washington University in St Louis School of Medicine.

Antisense oligonucleotides (ASOs) were purchased from Integrated DNA Technologies (IDT, Coralville, Iowa, USA) designed with 2′-O-methoxyethyl (2′-MOE) sugar ring and phosphorothioate (PS) backbone modifications. The 2′-MOE modification is well-tolerated, enhances target affinity, and increases resistance to degradation by nucleases while the PS modification enhances stability and cellular uptake (Schoch and Miller, 2017). Tau lowering ASOs were designed using a “gapmer” strategy, which consists of 2′-MOE modifications flanking the central region of unmodified nucleotides. Control ASOs with same modifications but without a specific target were included in select experiments. Synthesized ASOs were subjected to standard desalting procedure to remove small organic contaminants and high performance liquid chromatography (HPLC) purification to separate full-length oligonucleotides from truncated species based on hydrophobicity or charge difference. Sodium salt exchange was then applied on these ASOs to remove triethylammonium acetate residue from HPLC purification. ASO names and sequences are listed in the Supplementary Table 1. Control and tau-targeting sequences were identical in both sequence and chemistry to previously published studies in C57BL/6 (DeVos et al., 2013), hTau (Self et al., 2018), and PS19 mice (DeVos et al., 2017).

C57BL/6, hTau, and PS19 mice were anesthetized with isoflurane via continuous inhalation and placed into a stereotaxic frame during surgical procedures. For ASO infusion into the lateral ventricle of C57BL/6 and PS19 mice, 28-day ALZET osmotic pumps, which deliver ASOs at a rate of 6 μl/day (equivalent to 30 μg/day) (Durect, Model 2004) were implanted as previously described (DeVos and Miller, 2013b). hTau mice received a single intracerebroventricular (ICV) bolus injection (300 μg) for a treatment duration of 1 month as previously described (DeVos and Miller, 2013b). For euthanasia and tissue collection, mice were anesthetized with isoflurane and perfused with ice-cold 1X phosphate buffered saline (PBS). Brains were collected rapidly, and the left hemisphere (contralateral to the catheter placement or ICV injection) was drop-fixed into ice-cold 4% paraformaldehyde for histological processing. The right hemisphere (ipsilateral to the catheter placement or ICV injection) was dissected into separate pieces for RNA and protein analyses, snap frozen in liquid nitrogen, and stored at −80^°C until use.

Total RNA was extracted from right hemisphere brain lysates using a QIAGEN RNeasy kit (Catalogue #74104, QIAGEN, Hilden, Germany) per manufacturer’s instructions. Gene targets were amplified using Express One-Step Superscript qRT-PCR universal kit (Catalogue #11781200, ThermoFisher Scientific) and measured on a QuantStudio 12K Flex Real-Time PCR system (ThermoFisher Scientific). Target mRNA expression levels were normalized to mouse Gapdh mRNA levels and calculated using the ΔΔCt method. Primer and probe sequences were as follows: mouse Mapt: forward 5′-GAACCACCAAAATCCGGAGA-3′, reverse 5′-CTCT TACTAGCTGATGGTGAC-3′, probe 5′-/56-FAM/CCAAGAAGG TGGCAGTGGTCC/3IABkFQ/-3′; human Mapt: forward 5′-AG AAGCAGGCATTGGAGAC-3′, reverse 5′-TCTTCGTTTTACC ATCAGCC-3′, probe 5′-/56-FAM/ACGGGACTGGAAGCGATGA CAAAA/3IABkFQ/-3′; Gapdh: forward 5′-TGCCCCCATG TTGTGATG-3′, reverse 5′-TGTGGTCATGAGCCCTTCC-3′, probe 5′-/56-FAM/AATGCATCCTGCACCACCAACTGCTT/3IA BkFQ/-3′.

Total protein was extracted from tissue sections from the right hemisphere of mouse brain using RAB buffer (100 mM MES, 1 mM EDTA, 0.5 mM MgSO₄, 750 mM NaCl, 20 mM NaF, 1 mM Na₃VO₄) containing phosphatase and protease inhibitors and homogenized with a handheld tissue homogenizer. Bicinchoninic Acid (BCA) assay was used to quantify extracted protein.

For quantification of human and mouse tau protein, sandwich ELISA of tau-5 coating antibody (20 μg/ml; Millipore Cat# 577801-100UG RRID:AB_212534) with either biotin-conjugated HT-7 (for human tau, 0.3 μg/ml; ThermoFisher Scientific Cat# MN1000B RRID:AB_223453) or BT-2 (for mouse tau, 0.3 μg/ml; ThermoFisher Scientific Cat# MN1010B RRID:AB_10974155) antibodies was performed. Initially, 96-half-well plates were coated and incubated overnight with tau-5 antibody diluted in carbonate coating buffer (0.2M NaHCO₃/Na₂CO₃). The plates were then blocked with 4% bovine serum albumin (BSA)/PBS at 37^°C. Standards of recombinant human tau (2N4R; rPeptide, Bogart, GA) or mouse tau (mTau40, 432aa, produced in the laboratory of Eva-Maria Mandelkow) and samples were diluted in sample buffer (0.25% BSA, 300 mM Tris, and 1X protease inhibitor cocktail in PBS), loaded on plates, and incubated overnight at 4^°C. HT7 and BT-2 capture antibodies followed by streptavidin-poly-HRP-40 conjugate (1:4000; Fitzgerald, Acton, MA) were applied and detected by the addition of 3,3′,5,5′-tetramethylbenzidine liquid substrate, Super Slow (T5569, Sigma) reagent. The plates were read using a BioTek microplate reader (Biotek Epoch, Winooski, VT) at 650nM.

The left hemisphere of mouse brains remained in 4% paraformaldehyde for 24 h, was transferred to a 30% sucrose solution for cryoprotection, and frozen in cold (−20 to −35°C) 2-methylbutane. Frozen brains were sectioned into 40 μm coronal sections using a microtome and stored free-floating in a cryoprotectant solution (30% ethylene glycol, 30% glycerol, 10% 0.2 M phosphate buffer, in water) at −20°C until use.

For immunofluorescence, three sections per brain (approximate bregma level range −1.75 to −3.00 mm) were selected from saline- and ASO-treated PS19 mice and non-transgenic littermates. All immunofluorescence procedures were performed at room temperature except primary antibody incubation. Sections were incubated in a Tris-buffered saline (TBS) −0.1% Triton X-100 blocking solution containing 2% normal donkey serum, 4% BSA, 0.1% hydrogen peroxide, and 100mM glycine and then incubated in primary antibody at 4^°C overnight (Pan ASO 1:1000, gift from Ionis Pharmaceuticals, Carlsbad, CA, USA; NeuN 1:500, 266004 Synaptic Systems, Goettingen, Germany; GFAP 1:1000, AB5541 Millipore-Sigma; Iba1 1:500 ab5076 Abcam, Cambridge, United Kingdom). Sections were incubated in blocking solution prior to incubation in secondary antibodies (all at 1:1000; for ASO: anti-rabbit Alexa Fluor Plus 647, A32795 ThermoFisher Scientific; for NeuN: anti-guinea pig Alexa Fluor 555, A21435 Invitrogen or anti-chicken Alexa Fluor 594, 703-585-155 Jackson Immuno Research Labs, West Grove, PA, USA; for GFAP: anti-chicken Alexa Fluor 488, 703-545-155 Jackson Immuno Research Labs; for Iba1: anti-goat Alexa Fluor 488, 705-545-147 Jackson Immuno Research Labs). Sections were washed, incubated in DAPI (1:1000), and mounted on slides. Slides were covered with coverslips using Fluoromount-G mounting media (SouthernBiotech, Birmingham, AL, USA) and sealed with nail polish. Images were collected from the CA3 region of the hippocampus at bregma −1.85 mm using a Zeiss LSM 880 microscope (Zeiss, Oberkochen, Germany) with a 40 × 1.3NA oil objective and the Airyscan super-resolution detector with tile scanning. Following initial image acquisition, all images were Airyscan Processed using Zeiss Zen Blue 3.6. Overview images had an XY resolution of 70nm per pixel and are presented as a maximal intensity projections of a 1.2μm z-stack. For cell-specific imaging, a 10.2μm z-stack was collected with a resolution of 50nm in XY and 300nm in Z. Maximal intensity projections are shown for the XY plane of each identified cell to show clear cellular morphology and 1.8μm slices are shown for the XZ and YZ orthogonal projections to identify ASO that is localized within the specific cell types.

For immunohistochemistry of phosphorylated tau, six sections per brain (approximate bregma level range −1.65 to −2.98 mm) were selected from non-transgenic and saline- or ASO-treated PS19 mouse tissue. All immunohistochemical procedures were performed at the room temperature except primary antibody incubation. Sections were incubated in 0.3% hydrogen peroxide prior to blocking in 3% non-fat dry milk in PBS-0.25% Triton X-100 solution. Sections were then incubated in primary antibody at 4^°C overnight (biotinylated AT8, 1:500, MN1020B, ThermoFisher Scientific). The biotin signal was amplified using an ABC Elite kit (1:400; PK-6100, Vector Laboratories, Burlingame, CA, USA) and detected by reaction with 3, 3‘-diaminobenzidine (DAB) solution (SK-4100, Vector Laboratories). Sections were then washed and mounted on slides. Mounted sections were dehydrated through an ethanol gradient followed by incubation in xylene. Slides were covered with coverslips using mounting media (Cytoseal 60, #8310-4, ThermoFisher Scientific) and imaged at 20x on a Hamamatsu NanoZoomer HT whole slide imager (Hamamatsu Photonics, Japan).

To quantify phosphorylated tau, section images from AT8 immunohistochemistry were exported at 2.5x (cortex) or 5x (hippocampus) magnification using the NDP view (Hamamatsu) program and analyzed in ImageJ as previously described (Schoch et al., 2016). In brief, cortical and hippocampal regions across six sections were outlined as ROIs and a uniform, global threshold used to identify AT8 + reactivity within each region. The amount of AT8 + reactivity was recorded as a percentage of the total ROI area.

To quantify gliosis, images from saline- and ASO-treated PS19 and non-transgenic littermate controls stained for Iba1 (microglia) and GFAP (astrocytes) were exported using ZEN 3.4 software (ZEISS) and analyzed in ImageJ. Hippocampal regions across three sections were outlined as ROIs and a uniform, global threshold used to identify Iba1 + or GFAP + reactivity within each region. The amount of reactivity was recorded as a percentage of the total ROI area.

Data were represented in mean ± SEM and analyzed with GraphPad Prism (version 9). mRNA levels and protein expression were analyzed using unpaired t-test (for two groups) or ordinary one-way ANOVA with Tukey’s post-hoc test (for three or more groups). Analysis of immunohistochemical data (i.e., phosphorylated tau and gliosis) was done using one-way ANOVA with Tukey’s post-hoc test. A p value of < 0.05 was deemed significant.

We have previously demonstrated that tau-targeted ASOs reliably distribute throughout the brain and can be taken up by neurons and glia in the central nervous system (CNS) (DeVos et al., 2013, 2017). To test the ability of a purchased ASO to distribute and localize to cell types within the CNS, we performed immunofluorescence to visualize localization of the ASO in neurons, microglia, and astrocytes. We noted co-localization of the ASO with NeuN (Figures 1A3, B3), Iba1 (Figures 1B1, B2), and GFAP (Figures 1A1, A2), confirming neuronal and glial uptake of ASO.

Although mouse tau does not form pathological tau aggregates in wildtype mice, mouse tau may be used as a target to inform upon the physiological function of tau within the CNS. To test the effect of a purchased mouse tau-targeted ASO, we administered mouse tau knockdown (mTau KD) ASO via intraventricular osmotic pump to C57BL/6 mice and measured mTau mRNA expression 2 months after pump implantation. C57BL/6 mice treated with mTau KD ASO showed significant reduction in the levels of mTau mRNA expression levels with 73% and 68% decrease (Figure 2A) relative to saline or control ASO treated mice, respectively. No significant difference in mTau mRNA expression was observed between saline and control ASO treated mice (P = 0.1265). We also measured changes in the protein expression of mTau protein after ASO administration by ELISA. We found reduction in the levels of mTau protein in mice treated with mTau KD ASO with 61% and 56% decrease (Figure 2B) compared to saline or control ASO treated mice, respectively.

Human tau mice (hTau mice) express all six isoforms of human tau under control of the human tau promoter in the absence of endogenous mouse tau (Andorfer et al., 2003), offering a model that closely mirrors human tau expression. Our previous study in hTau mice treated with hTau KD ASO showed significant reduction in human tau mRNA and protein levels (Self et al., 2018). To test the effect of the purchased hTau KD ASO, we administered hTau KD ASO 1 via single ICV injection (similar treatment paradigm to our previous study (Self et al., 2018)) to hTau mice and measured hTau mRNA and protein levels. hTau KD ASO 1 treatment afforded a significant reduction (84% decrease) in the levels of human tau mRNA expression levels relative to control ASO (Figure 3A). We also measured human tau protein expression using ELISA after hTau KD ASO 1 administration in hTau mice. ASO-mediated hTau knockdown in hTau mice showed a significant reduction (62% decreases) in human tau protein expression relative to control ASO treated hTau mice (Figure 3B).

Mutant tau expression in PS19 closely recapitulates the time course of tau phosphorylation and aggregation, neuroinflammation, and cell loss evident in human tauopathies (Yoshiyama et al., 2007). In our past studies, we demonstrated the ability of hTau-lowering ASOs to reverse tau pathology and extend survival in PS19 mice (DeVos et al., 2017). Using a purchased version of the ASO, we administered hTau KD ASO 2 via osmotic pump to PS19 mice and measured human tau mRNA and protein expression 2 months after ASO administration. Non-transgenic littermates treated with saline were included as treatment controls. hTau KD ASO 2 treated mice showed a significant reduction in the levels of human tau mRNA (31% decrease) relative to saline treated mice without a change in mouse tau mRNA expression, confirming the specificity of the human tau target (Figures 3C, D). We observed reduced human tau protein expression in PS19 mice treated with hTau KD ASO (54% decrease) relative to saline treated PS19 mice (Figure 3E).

PS19 mice develop phosphorylated tau at 6 months of age, accompanied by significant microgliosis beginning at 3 months age and astrogliosis at 6 months age (Yoshiyama et al., 2007). To investigate pathological outcomes in PS19 mice after hTau lowering with a purchased ASO, we performed immunohistochemistry using the AT8 antibody to identify phosphorylated tau reactivity within the cortex and hippocampus (Figures 4A–D). PS19 mice treated with hTau KD ASO 2 exhibited reduced phosphorylated tau staining compared to saline treated PS19 mice in both the cortex (49% decrease) and hippocampus (39% decrease), although not statistically significant (Figures 4E, F). Next, to test the effect of hTau KD ASO 2 on microglial and astrocytic activation, we probed mouse brain sections with antibodies to Iba1 (Figures 4G, H) and GFAP (Figures 4I, J). PS19 mice treated with hTau KD ASO 2 showed significant reductions in Iba1 (81% decrease) and GFAP (45% decrease) relative to saline treated mice (Figures 4K, L).