Hyperphosphorylated Human Tau Accumulates at the Synapse, Localizing on Synaptic Mitochondrial Outer Membranes and Disrupting Respiration in a Mouse Model of Tauopathy

Abstract

Neurogenerative disorders, such as Alzheimer's disease (AD), represent a growing public health challenge in aging societies. Tauopathies, a subset of neurodegenerative disorders that includes AD, are characterized by accumulation of fibrillar and hyperphosphorylated forms of microtubule-associated protein tau with coincident mitochondrial abnormalities and neuronal dysfunction. Although, in vitro, tau impairs axonal transport altering mitochondrial distribution, clear in vivo mechanisms associating tau and mitochondrial dysfunction remain obscure. Herein, we investigated the effects of human tau on brain mitochondria in vivo using transgenic htau mice at ages preceding and coinciding with onset of tauopathy. Subcellular proteomics combined with bioenergetic assessment revealed pathologic forms of tau preferentially associate with synaptic over non-synaptic mitochondria coinciding with changes in bioenergetics, reminiscent of an aged synaptic mitochondrial phenotype in wild-type mice. While mitochondrial content was unaltered, mitochondrial maximal respiration was impaired in synaptosomes from htau mice. Further, mitochondria-associated tau was determined to be outer membrane-associated using the trypsin protection assay and carbonate extraction. These findings reveal non-mutant human tau accumulation at the synapse has deleterious effects on mitochondria, which likely contributes to synaptic dysfunction observed in the context of tauopathy.

Keywords: Alzheimer’s disease; aging; bioenergetics; phosphorylation; proteomics; synaptic mitochondria; tau; tauopathy.

FIGURE 1

Phosphotau protein accumulates in synaptosomes of htau mice at 8-months of age. Synaptosomes were from 8-month-old WT and htau mice by discontinuous Percoll gradient, lysed, and evaluated by immunoblot for expression of total tau and phospho-tau. (A) Representative immunoblot results for male and female animals. (B) Densitometric quantitation is shown for males (blue) and females (red). Relative protein expression was determined by normalizing the indicated tau isotype to synaptophysin. Student’s T-tests were performed to determine statistical significance between WT and htau (*p < 0.05, ^**p < 0.01; n = 6 per genotype, males and females were combined for analysis). ns, non-significant.

FIGURE 2

Mitochondrial bioenergetics are impaired in synaptosomes from htau mice. Synaptosomes were isolated from male and female WT and htau mice at 8-month of age and bioenergetics were assessed using the Seahorse MitoStress Test. Oxygen consumption rates (A) line graph and (B) bar plot. (C) Quantification of respiration profiles for basal, ATP-linked, proton leak, maximal, spare, and non-mitochondrial respiration. (D) Quantification of ATP production assay under identical substrate conditions with or without the addition of oligomycin. Tukey’s fence outlier test was used to exclude outlier biological replicates. Two-way ANOVA with Sidak’s multiple comparisons test was used to determine statistical significance (*p < 0.05, ^**p < 0.01; WT male n = 6, htau male n = 5, WT female n = 5, htau female n = 5; with 5 technical replicate wells per biological replicate).

FIGURE 3

Tau expression alters synaptic mitochondrial respiration. Synaptic and non-synaptic mitochondria were isolated by discontinuous Percoll density gradient centrifugation from 5- and 8-month-old WT and htau mice and complex II-driven mitochondrial respiration (A–D) was assessed using a Seahorse XFe24 Analyzer with the coupling assay. Seahorse quantifications were normalized to protein content on a per well basis. Statistical significance was determined by 2-way ANOVA with Sidak’s multiple comparisons test (**p < 0.001, ****p < 0.0001; n = 3, with 3 or 4 technical replicate wells).

FIGURE 4

Canonical pathway enrichment analysis. Heatmap representing enriched metabolic pathways produced by IPA canonical pathway analysis of the SWATH-MS Log₂ htau/WT protein expression data generated from an all-vs.-all comparison of the mitochondrial populations. Score filter p-value cutoff = 1.3 (log10) and z-score cutoff = 1 (absolute value). Blue blocks and red blocks represent inhibited and activated metabolic pathways, respectively. Gray dots represent metabolic pathways that did not achieve significance in the given experimental comparison.

FIGURE 5

Identification of differentially expressed mitochondrial proteins. (A) The distribution of p-values (−log10) and z-scores (log₂) in 1,313 quantified mitochondrial proteins between the htau and WT mice. A total of 65 (5 months, synaptic), 47 (8 months, synaptic), 38 (5 months, non-synaptic), and 49 (8 months, non-synaptic) proteins were selected as differentially expressed, which exhibited a p-value < 0.05 and a z-score > 1 standard deviation (highlighted in red). (B) Quantification of relative expression of indicated differentially expressed proteins as determined by SWATH-MS analysis. Differential expression (p-value < 0.05 and z-score > 1 standard deviation (annotated by a*).

FIGURE 6

Mitochondrial content of synaptosomes is unaffected by human tau expression. Synaptosomes were isolated by discontinuous Percoll gradient from WT and htau mice at 8-months of age, doubled-labeled with FM 1-43 (a marker of synaptic membranes) and MitoTracker Deep Red (MTDR) and analyzed by flow cytometry. (A) Percent of double-labeled synaptosomes. (B) Median fluorescence intensity of each label. (C) Immunoblot analysis of isolated synaptosomes comparing Vdac1 (mitochondrial marker, green) and synaptophysin (synaptosome marker; red) and (D) densitometric quantification. Individual biological replicate points are included in violin plots male values in blue and female in red. Two-way ANOVA was used to assess statistical significance for flow experiments (n = 6 per genotype, males and females combined) and two-tailed unpaired Student’s T-Test for immunoblot (n = 6 per genotype, males and females combined). ns, non-significant.

FIGURE 7

Tau accumulation does not promote mitochondrial genome instability. mtDNA was isolated from purified synaptosomes and analyzed by nested PCR to identify deletions and real-time quantitative PCR to determine mtDNA copy number. (A) Representative agarose gel images of nested PCR, deletion events are indicated by the presence of banding between ∼900 and ∼200 bp. The 12S region was amplified in parallel as loading control and to represent a region of mtDNA that is not susceptible to deletions. (B) Amplification curves of mtDNA copy number obtained by RT-qPCR with a TaqMan probe designed against 3155-3309 of the mouse mitochondrial genome. (C) Threshold cycle time values for mtDNA copy number. In B and C values were normalized to mg of synaptosome input for DNA isolation. Statistical significance was tested by two-tailed unpaired Student’s T-Test [n = 6 per sex and genotype; males (blue) and females (red) combined].

FIGURE 8

Total and phosphorylated tau is elevated in synaptic mitochondrial isolates from htau mice. Synaptic mitochondria were isolated by discontinuous Percoll gradient from 5- and 8-month-old WT or htau mice, solubilized with SDS buffer and analyzed by immunoblot for (A) total and (B–D) phosphoforms of tau (green). Quantification of immunoblots was done by densitometry, normalizing the indicated tau form to the mitochondrial marker Vdac1 (red). Statistical significance was determined by multiple T-Tests using the two-stage setup method (Benjamini, Krieger, and Yekutieli) (*p < 0.05, ^***p < 0.001, n = 3 per genotype, males).

FIGURE 9

Mitochondrial associated tau is not membrane integrated. (A,B) Mitochondria isolated from 8-month-old htau mice were incubated on ice in a sodium carbonate extraction buffer and fractionated by ultracentrifugation to differentiate membrane associated from membrane-integrated proteins. The resulting fractions were assessed by immunoblot for the indicated proteins (P = pellet/insoluble (membrane-integrated), S = soluble (membrane associated). (A) Blot for Hsp60 and OxPhos as representative controls for the assay. (B) Blot of total tau. (C) Synaptic mitochondria isolated from 8-month-old htau mice were digested with the indicated concentration of trypsin on ice for 30 min, and then solubilized with SDS buffer and analyzed by immunoblot for total Tau, Sdha, and Vdac1 to determine the extent of protease digestion. (D) Diagram illustrating mitochondrial sub-localization of the proteins probed for in (A–C). Red asterisks on Vdac1, Tau, and Sdha represent relative epitope location for the antibodies used in immunoblotting following protease treatment.