Vascular tau in Alzheimer's and other tauopathies
Soluble tau aggregate accumulation in brain microvasculature of P301S(PS19) mice is associated with profound deficits in endothelium-dependent vasodilation. (A) Impaired endothelium-dependent cerebral blood flow (CBF) responses in P301S(PS19) mice in response to topical acetylcholine (ACh) stimulation as compared to WT animals worsen with age (post hoc comparisons to age-matched WT group: * indicates Tukey’s q(126)=3.80, ** indicates Tukey’s q(126)>4.66, P<0.007; **** indicates Tukey’s q(126)>6.16, P<0.0001; bracket highlights significant worsening of endothelial dysfunction with age in P301S(PS19) mice, ** Tukey’s q(126)=4.61). Data (means ±SEM of n=6-7 per group) are plotted against ACh-induced vasodilation in C57BL/6 mice, not included in the analysis. (B) Soluble tau aggregates accumulate in brain microvasculature of P301S(PS19) mice. Representative images of brain sections from P301S(PS19) and WT controls showing soluble tau aggregate (T22, green) and total tau (Tau 5, cyan) immunoreactivity in lectin-stained microvasculature (red); (C-D) Accumulation of tau in brain microvasculature isolated from P301S(PS19) mice. C, Representative electropherograms from capillary electrophoresis immunoassays showing increased tau in brain vasculature isolated from P301S(PS19) mice. D, Quantitative analyses of data in (C)(t(5.064)=3.116, *p<0.03, n=6 per group). Data are representative images and means ± SEM.
Heparan sulfate proteoglycan (HSPG)-dependent internalization of soluble extracellular tau aggregates by brain microvascular endothelial cells triggers endogenous tau phosphorylation (Thr231). (A-C) Soluble aggregated tau internalization triggers endogenous tau phosphorylation in primary human brain microvascular endothelial cells (HBEC). (A) Representative images of primary HBEC exposed to soluble aggregates of V5-tagged recombinant human tau-441 or vehicle (control) in the presence or absence of heparin, immunostained with antibodies for V5 (green) and α/β-tubulin (red), and counterstained with DAPI (blue). (B) Representative electropherograms from capillary electrophoresis immunoassays for phosphorylated tau (T231) and β-actin in lysates from HBEC treated with recombinant human cytokeratin-8 (KRT8), monomeric tau protein (M. Tau), unlabeled soluble tau aggregates (O.Tau), or vehicle (control). (C) Quantitative analyses of data in B (1-way ANOVA, F(3,22)=3.115, P=0.047, *, n=4-12). (D-F) Tau is expressed in primary microvascular endothelial cells. (D) Quantitative real-time PCR measurements of tau mRNA abundance in human SK-N-SH neuroblastoma cells, HBEC, and primary human umbilical vein endothelial cells (HUVEC) show comparable expression of tau in cells of endothelial and neuronal origin F (2, 9) = 2.467, p=0.14, n=3 per group). (E) Traditional Western blot showing tau protein content in HBEC with β-actin as reference. (F) Tau immunoreactivity in HBEC (Tau5, green) is found in close proximity to tubulin (α/β-tubulin, red). (G-H) Atomic force microscopy profiling of tau. (G) Representative images of soluble tau aggregates measured by atomic force microscopy (AFM). (H) Size distribution of tau particles as measured by AFM show overlapping distributions of soluble tau aggregates (>12 nm in length) in non-tagged and V5-tagged soluble tau aggregate preparations. Data are representative images, AFM profiles, and means ± SEM.
Removal of soluble tau aggregates from brain ameliorates brain microvascular deficits in P301S(PS19) mice. (A-B) Treatment with TOMA ameliorates profound deficits in endothelium-dependent vascular responses in somatosensory cortex of P301S(PS19) mice, restoring vascular responses to levels comparable to those of 4 month-old animals. A, ACh-stimulated vascular responses; B, Quantitative analyses of 25 min data in A. Interaction of time and treatment group by 2WR ANOVA: F(10,50)=4.91, P<0.0001; difference from WT by Holm-Sidak’s post hoc indicated as * t(60)=2.86, p=0.02; ** t(60)=3.33, p=0.003; *** t(60)=4.04, p=0.0005; **** t(60)=5.31, p<0.0001; # indicates difference from PS19+TOMA, for all #s, t(60)>2.34, p<0.046 by Holm-Sidak’s post hoc; n=4-5 mice/group. Data from 4-month-old WT and P301S(PS19) groups serve as reference and were not included in the analysis. (C-D) Treatment with TOMA reduces levels of soluble tau oaggregates (tau oligomers) in brains of PS19(P301S) mice. C, Representative immunoblots of cortical lysates of P301S(PS19) mice treated with isotype-matched IgG or TOMA; D, Quantitative analyses of data in C (F (2, 19) = 6.807, by 1-way ANOVA, *p<0.05, **p<0.01 by Tukey’s post hoc test, n=7-8/mice per group). (E-F) TOMA treatment reduces levels of vascular tau in brains of P301S(PS19) mice. Vascular tau is increased in brains of P301S(PS19) mice treated with non-specific isotype-matched IgG, but not in brains of P301S(PS19) mice treated with TOMA. E, Representative images (100X) of tau immunoreactivity (green) associated with lectin-reactive brain vasculature (red), counterstained with DAPI (blue) staining. Scale bar is 50 µm. F, Quantitative analyses of data in E (H=6.839, P=0.026, *, Kruskal-Wallis test, n.s., no significant difference in experimental means for WT and PS19+TOMA groups, Dunn’s multiple comparison test, n=5-8). (G-H) Cortical brain vascular density in P301S(PS19) mice is unchanged by tauopathy or by TOMA immunotherapy. G, Representative images (40X) of lectin-reactive cortical brain vasculature (red) of P301S(PS19) animals treated with TOMA or isotype-matched IgG. Scale bar is 250 µm. H, Quantitative analyses of data in G (H=1.787, P=0.426, Kruskal-Wallis test). Data are representative images and means ± SEM.