The coding sequences for EGFP, TauA152T (C-terminal V5 tag), and TauP301L (C-terminal V5 tag) were subcloned into an AAV vector and sequence-verified using ABI3730 with Big Dye chemistry (Applied Biosystems, Foster City, CA). AAV vectors containing EGFP, TauA152T, and TauP301L under the control of the cytomegalovirus enhancer/chicken β-actin promoter, as well as a woodchuck post-transcriptional regulatory element and the bovine growth hormone polyA, were cotransfected with AAV helper plasmids into HEK293T cells. Cells were harvested and lysed in the presence of 0.5% sodium deoxycholate and 50 U/mL Benzonase (Sigma, St. Louis, MO) by freeze thawing 48 h post-transfection, and the virus was isolated using a discontinuous iodixanol gradient. Quantitative PCR was used to measure the genomic titer of each virus.

All animal procedures were approved by the Mayo Institutional Animal Care and Use Committee (IACUC) and are in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publications No. 80-23, revised 1996). Intracerebroventricular (ICV) injections with either TauA152T-AAV, TauP301L-AAV or EGFP-AAV were performed into mouse pups on postnatal day 0 (2.7E + 10 viral particles/ventricle; 2 μL/ventricle) as described (Chakrabarty et al., 2013). Briefly, newborn mice were cryoanesthetized and placed on a cold metal plate. A 10 μL Hamilton syringe with a 30-gauge needle was used to pierce the skull just posterior to bregma and 2 mm lateral to the midline, and 2 μL of AAV was injected into each of the lateral ventricles. Mice were euthanized by CO₂ inhalation at 3 months of age. The brain was immediately removed and frozen for RNA extraction and analysis. Please see Supplementary Table S1 for sample information.

Half brains were weighed and homogenized in 5× volume of TE buffer [50 mM Tris base (pH 7.4), 50 mM NaCl, 1 mM EDTA, 1 mM PMSF, 1× protease and phosphatase inhibitors]. For RNA isolation, 70 μL of homogenate was added to 210 μL Trizol LS (Life Technologies, Carlsbad, CA), and the mixture frozen at −80°C until extraction. Total RNA was isolated from brain tissue using the Direct-zol RNA Miniprep kit (Zymo Research, Irvine, CA) according to manufacturer’s instructions with in-column DNase I treatment. RNA concentrations and A260/280 ratios were measured using a Nanodrop.

The nCounter mouse glial profiling panel (NanoString, USA) was used to quantify expression in mouse RNA samples, with data analyzed using an nCounter FLEX Analysis System (NanoString, USA). Raw data were imported into nSolver 4.0 and analyzed with the nCounter advanced analysis paired with R (version 3.3.2). Following normalization and quality control, data was analyzed to assess differential expression, as well as perform pathway and cell type profiling analyses. Mice were age-matched and all samples were included on one cartridge, eliminating the need to control for age or batch effects. To determine whether differences between GFP-AAV, P301L-AAV, and A152T-AAV animals were statistically significant, differences between groups were assessed using 1-way ANOVA followed by a Tukey’s post-hoc test for multiple comparisons. All statistical analyses were performed in R and GraphPad Prism, and data are presented as mean ± SEM, with p < 0.05 considered statistically significant.

We previously observed that mice expressing the P301L mutation exhibit accumulation of insoluble hyperphosphorylated tau and microgliosis in the absence of neuronal loss (Cook et al., 2015; Carlomagno et al., 2019). Conversely, accumulation of soluble hyperphosphorylated tau, astrocytosis and neuronal loss are characteristic of mice expressing A152T tau (Carlomagno et al., 2019). To illuminate how these different patterns of tau deposition differentially impact glial cell biology, we performed a comprehensive analysis using the nCounter glial profiling panel to simultaneously quantify mRNA expression of 770 genes (Supplementary Table S2). Differential expression analysis revealed that P301L-AAV (Figures 1A,B) and A152T-AAV mice (Figures 1C,D) exhibit distinct changes compared to the control group (GFP-AAV).

In P301L-AAV mice, the expression of 23 genes was upregulated (Figures 1A,B) and four downregulated (Figures 1A,B). Among the top 10 upregulated genes were multiple markers associated with inflammatory microglia, including Clec7a, Cst7, Gpr84, and Cd84, suggesting a strong microglial response associated with insoluble tau deposition in P301L-AAV mice. In addition, genes linked with astrocyte activation (Gfap, Serpina3n) and inflammatory signaling (Ccl3, Nlrp3, Sh2d1a) were also significantly upregulated in P301L-AAV mice, along with Mapt itself. Downregulated genes include the immune response regulator Klrb1, genes involved in interferon signaling (Iigp1 and Nkg7), and Fgf2 involved in cellular growth signaling.

In A152T-AAV mice, the expression of 24 genes was upregulated and four downregulated compared to the control group (Figures 1C,D). Among the top upregulated genes, similar to P301L-AAV mice we observed increased expression of microglial activation markers (Clec7a, Lgals3, Gpnmb, Cd84) and the complement component C4a/b, in addition to astrocyte-related genes (Gfap, Serpina3n) in A152T-AAV mice. Additional astrocyte markers that were unchanged in P301L-AAV mice were increased in A152T-AAV mice, including Gdpd2, Ggta1, Aqp4, and Fbln5. Downregulated genes in A152T-AAV mice include genes involved in T cell signaling and cytokine responses (Il4), glycolipid metabolism (B3gnt5), phagocytosis and neurotrophin signaling (Prkcd), calcium signaling (Itpr1), and immune cell function (Cd163).

While both models showed increased Mapt expression (providing experimental validation of the panel) as well as upregulation of several common genes involved in neuroinflammation, including Clec7a, Gfap, Serpina3n, Cd84, and C4a/b, each tau mutation induced unique gene expression patterns (Figure 1E). In particular, an upregulation of genes associated with inflammatory microglial responses (Ccl3, Nlrp3, Gpr84, Sh2d1a), chemokines (Ccl6, Ccl3), and cytokine signaling (Il1a, Tmem88b) were only observed in P301L-AAV mice, suggesting a stronger microglial inflammatory phenotype. In contrast, A152T-AAV mice uniquely upregulated genes related to astrocyte function (Gdpd2, Ggta1, Aqp4, Fbln5), apoptotic signaling (Casp3), and homeostatic microglial functions (C1qb, Ctss), indicating a more prominent astrocytic response.

To better understand the biological significance of differentially expressed genes, we performed pathway-level analyses comparing both tau-AAV models to GFP-AAV controls (Figure 2A, Supplementary Table S3). Both models showed strong activation of multiple astrocyte-related pathways, including astrocyte differentiation/function and markers, consistent with the upregulation of glial fibrillary acidic protein (Gfap) and Serpin Family A Member 3N (Serpina3n). Similarly, microglial pathways, including the microglial neurodegenerative phenotype (MGnD) and microglial markers, showed marked increases in both P301L-AAV and A152T-AAV mice, reflecting the elevated expression of microglial genes like Clec7a, Cd84 and Cd68. While both models also exhibited decreased activity in antigen processing and presentation and interferon signaling pathways, the complement system and cytokine pathways were increased, aligning with the increased expression of inflammatory mediators like C4a/b. Moreover, consistent with the significant increase in Mapt expression in both models combined with tau’s known role in cytoskeletal regulation, cytoskeletal dynamics was one of the most upregulated pathways in both tau-AAV groups. Both models also showed activation of JAK-STAT signaling, while neurotrophin signaling and phagocytosis were most reduced in A152T-AAV mice.

In addition to examining pathway scores, we also performed cell type profiling to understand how distinct patterns of tau deposition may differentially impact cellular populations (Supplementary Table S4). Notably, astrocyte scores were significantly increased in both tau-AAV models compared to the GFP-AAV control group (Figure 2B), which may indicate that accumulation of hyperphosphorylated tau independent of changes in solubility may be linked with astrocytosis. Exhausted CD8 T cell scores were also significantly increased but only in P301L-AAV mice (Figure 2B), which suggests a relationship with insoluble tau accumulation. Finally, despite activation of both oligodendrocyte- and microglial-related pathways in both tau models (Figure 2A), oligodendrocyte and microglia cell type scores showed no significant differences between groups (Figure 2B).

To further illuminate functional differences between tau models, we performed pathway enrichment analysis, which revealed both common and model-specific changes. In particular, while both tau-AAV models showed enrichment in JAK-STAT signaling, hematopoietic cell lineage and cytokine-cytokine receptor interaction pathways were enriched only in P301L-AAV mice (Figure 2C), consistent with the elevated expression of Ccl3, Ccl6, and Cst7. Gene Ontology (GO) enrichment analysis revealed similar changes in both tau-AAV models, including significant enrichment in the biological processes (BP) positive regulation of superoxide anion generation’, “cell adhesion” and “innate immune response” (Figure 2D). However, additional terms were enriched only in P301L-AAV mice, including “neutrophil chemotaxis”, “inflammatory response”, “cellular response to lipopolysaccharide” and “positive regulation of tumor necrosis factor production”, collectively reflecting a pronounced inflammatory signature associated with insoluble tau deposition characteristic of the P301L-AAV model.

Both tau-AAV models were associated with significant differences in astrocyte differentiation and function compared to the GFP-AAV control group (Figure 3A), reflected in elevated astrocyte cell scores (Figure 2B). Although Sox9 and Aldh1l1 were increased in both P301L-AAV and A152T-AAV mice (Figure 3B) suggesting a more pronounced maturation profile (Sun et al., 2017; Yang et al., 2011), a key distinguishing pattern emerged in astrocyte activation markers. In particular, the classic reactive astrocyte markers Gfap and Serpina3n (Eng and Ghirnikar, 1994; Zamanian et al., 2012) exhibited higher expression in P301L-AAV mice (Figure 3B), implicating a relationship with insoluble tau accumulation. Moreover, while both models were associated with increased C4a/b expression, A152T-AAV mice were more prominently affected (Figure 3B).

Regarding microglia-associated impacts, despite the fact that overall microglial cell scores remained unchanged (Figure 2D), the microglial neurodegenerative phenotype (MGnD) was increased in both tau-AAV models compared to control (Figure 3C). Looking at individual targets, Axl and Ctsd were increased in both tau-AAV models (Figure 3D), which may implicate a compensatory upregulation in lysosomal activity to facilitate debris clearance and damage sensing capabilities (Kenessey et al., 1997; Kim et al., 2007; Fou et al., 2016). Of interest, Lgals3 (galectin-3) was only upregulated in A152T-AAV mice (Figure 3D), which is intriguing given that galectin-3 has been shown to bind hyperphosphorylated tau and augment tau aggregation (Siew et al., 2024). In addition, the increase in Clec7a and Cst7 expression was most robust in P301L-AAV mice, potentially indicative of a more pronounced inflammatory phenotype (Sala Frigerio et al., 2019; Keren-Shaul et al., 2017).

Oligodendrocyte differentiation and maturation was also increased in P301L-AAV mice compared to control (Figure 3E), with broad upregulation of several targets including Sox10, Mbp, Cd9, Plp1, and Tmem88b (Figure 3F), suggesting enhanced myelination potential (Deber and Reynolds, 1991; Galiano et al., 2006; Zeis et al., 2016; Rubinstein et al., 1996; Klugmann et al., 1997; Griffiths et al., 1998; Luders et al., 2017). While the oligodendrocyte lineage marker Sox10 was also increased in A152T-AAV mice relative to control, model-specific changes include a modest increase in Cnp and slight reduction in Zfp365 (Figure 3F), which may be indicative of subtle changes in oligodendrocyte differentiation patterns (Fan et al., 2023; Shimizu et al., 2014; Tohyama et al., 2015).

Our analysis revealed both common and model-specific changes in glucose and lipid metabolism, as well as mitochondrial function. In particular, changes in expression of genes associated with glucose metabolism were observed in A152T-AAV but not P301L-AAV mice (Figures 4A,B), including the simultaneous downregulation of both Dlat and Dld - key enzymes bridging glycolysis and the TCA cycle. Expression of the glucose transporter Slc2a5 (Glut5) was also decreased in A152T-AAV mice (Figure 4B). This coordinated suppression suggests a fundamental disruption in cellular energy production in the A152T-AAV model.

As lipid metabolism plays a crucial role in brain function and has been increasingly recognized as a key factor that is dysregulated in neurodegenerative diseases (Estes et al., 2021), it is intriguing that expression of Abca1, which is essential for lipid homeostasis in neurons and glia (Dean et al., 2001; Tachikawa et al., 2005; Hirsch-Reinshagen et al., 2004; Wellington et al., 2002), showed elevated expression in both P301L and A152T groups (Figure 4B). However, only A152T-AAV mice exhibited an increased Apoe and decreased B3gnt5 expression (Figure 4B), potentially implicating alterations in glycolipid metabolism in this model.

Building on alterations within metabolic pathways, we next examined expression of mitochondrial genes in both tau models (Figure 4C). Of note, only A152T-AAV mice exhibited a reduction in Atp5g3 expression (Figure 4C), which is essential for ATP synthesis during oxidative phosphorylation (Huang et al., 2013). In addition to Atp5g3, expression of other mitochondrial genes was also decreased in A152T but not P301L-AAV mice, including Chchd3 and Opa1 (Figure 4C). Of particular relevance, Chchd3 plays a critical role in maintaining mitochondrial crista architecture (Schauble et al., 2007; Dreger et al., 2005) and also regulates Opa1 (Darshi et al., 2011), which is a dynamin-related GTPase that is essential for mitochondrial fusion and cristae maintenance (MacVicar and Langer, 2016). Overall, these coordinated changes across glucose metabolism, lipid processing, and mitochondrial function pathways suggest A152T mutant tau expression is associated with widespread metabolic reprogramming that is distinct from the phenotype associated with P301L mutant tau expression.

In addition to metabolic alterations, A152T-AAV mice also exhibited more dramatic alterations than P301L-AAV mice across multiple cellular processes. In particular, apoptosis was significantly upregulated in A152T-AAV mice compared to both GFP-AAV and P301L-AAV groups (Figure 5A), which is likely driven by a robust increase in caspase 3 (Casp3) expression (Figure 5B). Increased expression of toll-like receptor 4 (Tlr4) in A152T-AAV mice was also trending but not significant, while levels of the nuclear import adaptor karyopherin-α1 (Kpna1) were significantly reduced by mutant A152T tau expression (Figure 5B). In contrast, levels of the apoptosis regulator bcl-2-like protein 4 (Bax) and poly (ADP-ribose) polymerase 2 (Parp2) expression were unchanged in either model.

Calcium signaling (Figures 5C,D) and phagocytosis pathways (Figures 5E,F) also showed mutation-specific effects, with alterations detected within the A152T-AAV but not P301L-AAV models. As key regulators of calcium signaling in the cell, expression of calmodulin 2 (Calm2), inositol 1,4,5-triphosphate receptor (Itpr1), and ATPase sarcoplasmic/endoplasmic reticulum Ca2+ transporting 2 (Atp2a2) were all reduced in A152T-AAV mice (Figure 5D). As the phagocytosis pathway is also significantly reduced in A152T animals (Figure 5E), it is noteworthy that several subunits of the V-ATPase complex (Atp6v1e1, Atp6v1g1, Atp6v1h) showed coordinated downregulation, implicating an impairment of lysosomal function (Colacurcio and Nixon, 2016). Additionally, reduced expression of protein kinase C delta (Prkcd) in A152T mice suggests mitophagy may also be perturbed (Munson et al., 2021). Collectively, these changes suggest that apoptotic signaling is elevated while cellular calcium homeostasis and phagocytosis/lysosomal function is inhibited by the accumulation of soluble hyperphosphorylated tau species in A152T mutant tau expressing mice.