Acetylation of tau inhibits its degradation and contributes to tauopathy

Abstract

Neurodegenerative tauopathies characterized by hyperphosphorylated tau include frontotemporal dementia and Parkinsonism linked to chromosome 17 (FTDP-17) and Alzheimer's disease (AD). Reducing tau levels improves cognitive function in mouse models of AD and FTDP-17, but the mechanisms regulating the turnover of pathogenic tau are unknown. We found that tau is acetylated and that tau acetylation prevents degradation of phosphorylated tau (p-tau). We generated two antibodies specific for acetylated tau and showed that tau acetylation is elevated in patients at early and moderate Braak stages of tauopathy. Histone acetyltransferase p300 was involved in tau acetylation and the class III protein deacetylase SIRT1 in deacetylation. Deleting SIRT1 enhanced levels of acetylated-tau and pathogenic forms of p-tau, probably by blocking proteasome-mediated degradation. Inhibiting p300 with a small molecule promoted tau deacetylation and eliminated p-tau associated with tauopathy. Modulating tau acetylation could be a new therapeutic strategy to reduce tau-mediated neurodegeneration.

Figure 1. Tau is Acetylated in Vitro and in Vivo

(A) Acetylation of h-tau (2N4R) by p300 but not pCAF under cell-free conditions, as shown by autoradiography. (B) MALDI-TOF spectrometry identified ac-lysines on h-tau by p300 in vitro. Red: lysines (K) with acetyl group. Underlined: sequence covered by MS analysis. Blue box: microtubule-binding domains. See Table-S1 for full list and Figure S1 for MS-MS spectra. (C–E) Ab708 specifically recognizes ac-tau. (C) Ab708 only recognized recombinant tau acetylated by GST-p300, not nonacetylated tau with GST alone. Similar levels of t-tau were detected with Ab707 and Tau 5 antibody. (D) Overexpressing p300 markedly enhanced ac-tau, detected with Ab708, in HEK293T cells. Levels of t-tau, detected with Tau 5, were similar with or without p300. Blots are representative of >5 experiments. (E) Putatively acetylated lysine sites recognized by Ab708. Ac-tau/t-tau levels in HEK293T cells expressing wildtype tau were set as 1. n = 4. *, P=0.012; ** P=0.003; ***, P=0.0003 (one-way ANOVA with Tukey-Kramer posthoc analysis). (F) Ab708 recognizes human ac-tau in brains of PS19 or hT-PAC-N transgenic mice, not in NTG littermates. Human t-tau was detected with Ab707 antibody; human and mouse t-tau was detected with Tau 5 antibody. See Figure S2 for the sequence similarity among human, mouse and rat tau. (G) Levels of Ab708-positive ac-tau were elevated in primary rat neurons as they matured in culture (DIV=5–12). n=2–7 from 2–3 independent experiments. ***, P<0.001 (DIV5 vs. DIV8 or DIV12); **, P<0.01 (DIV5 vs. DIV9–11). Values are means ± SEM (E, G).

Figure 2. Tau Is Acetylated by p300 Acetyltransferase

(A–C) Inhibiting p300, not pCAF, reduced ac-tau in HEK293T cells. (A) Inhibition of p300 or pCAF expression by siRNA transfections. Levels of p300/GAPDH or pCAF/GAPDH in control siRNA-transfected cells were set as 1. ***, P=0.0006 (p300 vs. control) or P<0.0001 (pCAF vs. control). (B) Representative western blots (from three experiments) showing levels of p300 or pCAF, ac-tau, t-tau, and GAPDH in cells transfected with control siRNA (CTRL) or siRNA targeting p300 or pCAF. (C) Inhibition of p300, not pCAF, reduced ac-tau levels. Levels of ac-tau/GAPDH or t-tau/GAPDH in control siRNA-transfected cells were set as 1. n=5–6. **, P=0.008 (paired t test). (D) Inhibiting p300 acutely with C646 (20 μM for 8 h) eliminated ac-tau without affecting t-tau levels in primary rat cortical neurons. Left: Representative western blot from three experiments. Right: Ac-tau/t-tau levels in vehicle-treated cells were set as 1. n=3. ***, P=0.0001 (unpaired t test). (E) Extended treatment with C646 (20 μM for 20 h) lowered t-tau in primary cortical neuron. Blots are representative of two experiments. Values are means ± SEM (A, C–E).

Figure 3. SIRT1 Deacetylates Tau in Culture

(A–D) SIRT1 overexpression lowered ac-tau levels in HEK293T cells. (A) Western blot showing expression of FLAG-tagged SIRT1, SIRT2, HDAC5, or HDAC6 with an anti-FLAG antibody. Blots are representative of 2–3 experiments. (B) Western blot showing levels of ac-tau, t-tau, tubulin, and ac-tubulin in cells overexpressing SIRT1, SIRT2, HDAC5, or HDAC6. Blots are epresentative of 2–3 experiments. (C) Overexpression of SIRT1, SIRT2 or HDAC6 significantly reduced levels of ac-tau/GAPDH. Levels of t-tau were also reduced by SIRT1 or HDAC6 overexpression. n=9–18 from 6–10 independent experiments. ***, P < 0.001 (Mock vs. SIRT1 or Mock vs. HDAC6); **, P < 0.01 (Mock vs. SIRT2) (two-way ANOVA and Bonferroni posthoc test). (D) Overexpression of SIRT1 significantly reduced ac-tau/t-tau. n=9–18 from 6–10 independent experiments, ***, P<0.001 (Mock vs. SIRT1) (one-way ANOVA and Tukey-Kramer posthoc test). (E–H) Inhibition of SIRT1 elevated ac-tau in HEK293T cells. (E) Inhibition of SIRT1, SIRT2, or HDAC6 expression mediated by siRNA transfections. n=4–6 from 2–3 experiments. **, P =0.0015; ***, P=0.0001 (SIRT2 vs. control) or P=0.001 (HDAC6 vs. control) (paired t test). (F) Western blot showing levels of ac-tau, t-tau, tubulin, and ac-tubulin in cells transfected with control siRNA or siRNA targeting SIRT1, SIRT2, or HDAC6. Blots are representative of 2–3 experiments. (G–H) Inhibition of SIRT1, significantly elevated levels of ac-tau/GAPDH (G) or ac-tau/t-tau (H). n=4–6 from 2–3 experiments. *, P<0.05 (paired t test). Levels of deacetylase/GAPDH (E), ac-tau or t-tau/GAPDH (G), and ac-tau/t-tau (H) in control siRNA-transfected cells were set as 1. Also see Figure S3 for comparison of levels of ac-tau in MEFs with (SIRT1^+/+) or without SIRT1 (SIRT1^−/−). (I) Deacetylation of Tau3KR by SIRT1. Left: Representative western blot showing levels of ac-tau, t-tau, FLAG-tagged SIRT1, and GAPDH. Right: Ac-tau/t-tau levels in mock-transfected cells expressing wildtype tau were set as 1. n=10–20 from 4–10 independent experiments. ***, P<0.001; ns, not significant (one-way ANOVA and Tukey-Kramer posthoc analysis). Values are means ± SEM (C–E, G–I)

Figure 4. SIRT1 Reduces Tau Acetylation in Neurons and in Vivo

(A) Western blot showing expression of SIRT1 in primary cortical neurons during maturation in culture (DIV5–11). Blots are representative of 2–3 independent cultures. (B) Levels of endogenous ac-tau relative to t-tau correlated negatively with levels of SIRT1 in primary rat neuronal cultures (DIV5–9). Levels of SIRT1 or ac-tau/t-tau at DIV=5 were set as 1. n=20 independent measurements. P=0.0007, Pearson correlation coefficient r²=0.4791. (C) Deleting SIRT1 in neurons elevated levels of ac-tau relative to t-tau. Neurons cultured from SIRT1^F/F mice were infected with control virus or virus expressing cre recombinase (Lenti-cre). Both cultures were also infected with a lentiviral vector expressing h-tau. n=8. P=0.001, (unpaired t test). (D) Acetyl-specific antibody (9AB) recognized tau acetylated by GST-p300, but not non-ac-tau. Also shown is the sequence of the antigen used to generate 9AB. (E) Overexpressing p300 enhanced 9AB-positive ac-tau in HEK293T cells. Levels of t-tau, detected with Tau 5, were similar with or without p300 overexpression. Blots are representative of three independent experiments. (F) Deletion of SIRT1 elevated ac-tau relative to t-tau in the brain. Left: Representative western blots showing levels of ac-tau, t-tau, and GAPDH. Right: Levels of ac-tau/t-tau in SIRT1^+/+, SIRT1^+/−, and SIRT1^−/− brains. n=3–6 mice/genotype. *, P=0.02 (SIRT1^+/− vs. SIRT1^−/−) (one-way ANOVA and Tukey-Kramer posthoc test). Values are means ± SEM (C, F)

Figure 5. SIRT1 Interacts with Tau

(A) SIRT1 directly deacetylated ac-tau in vitro. Ac-tau/t-tau levels in the absence of immunoprecipitated SIRT1 were set as 1. Ac-tau was detected with Ab708 antibody. n=5 from two experiments. **, P=0.0063 (unpaired t test). (B) GST pull-down assays. GST-tau protein (lanes 7–9) or GST alone (lanes 4–6) was incubated with lysates of cells transfected with FLAG-tagged SIRT1 or of nontransfected cells. Lanes 1, 4 and 7: 0.1% Triton X-100; lanes 2, 5, and 8: 0.5% Triton X-100; lanes 3, 6, and 9: 0.5% NP-40. Data shown are representative of two experiments. The lower band is likely to represent the cleavage product of SIRT1 observed previously (Ohsawa and Miura, 2006). (C) Coimmunoprecipitation assays. HEK293T cells were transfected with a plasmid encoding FLAG-tagged human tau. Cell lysates were collected 24 h later, immunoprecipitated with an anti-FLAG antibody, and immunoblotted with Tau 5 or an anti-SIRT1 antibody. Lanes 1–2: input; lane 3: no primary antibody; lane 4: anti-FLAG antibody. Values are means ± SEM (A).

Figure 6. Acetylation Slows Tau Turnover by Inhibiting Its Ubiquitination

(A) Inhibiting SIRT1 with EX527 (50 μM) elevated ac-tau and p-tau in rat primary neurons (DIV=10). Left: Representative western blots. p-tau was detected with AT8. Right: Levels of ac-tau/t-tau or p-tau/t-tau in vehicle-treated cells were set as 1. n=6 independent treatments. ***, P<0.001; *, P<0.05 (paired t test). (B) Deletion of SIRT1 elevated AT8-positive p-tau in the brain. n=3–4 mice/genotype. *, P<0.05 (SIRT1^+/+ vs. SIRT1^−/−) (one-way ANOVA and Tukey-Kramer posthoc test). (C and D) Tau3KR was more stable than wildtype tau in primary neurons. Cells were infected with Lenti-hTauwt or Lenti-hTau3KR and treated with CHX for 8–32 h 4 days after infection (DIV=9). (C) Representative western blot of 2 experiments showing ac-tau, t-tau, and GAPDH. (D) The turnover of t-tau was slower in cells expressing Tau3KR. t-tau/GAPDH levels in cells harvested at time 0 were set as 1. n=3–5 from 2 experiments. *, P=0.04 (8 h), P=0.015 (24 h); ***, P<0.0001 (32 h) (unpaired t test for each time-point). (E and F) SIRT1 inhibitor EX527 (1–50 μM) suppressed tau ubiquitination and elevated ac-tau in a dose-dependent manner. Blots are representative of 3 experiments. See Figure S4, which shows that resveratrol-enhanced tau ubiquitination depended on SIRT1's deacetylase activity. (G–K) The SIRT1 inhibitor EX527 increases the half-life of tau in rat primary neurons (DIV=8) in a dose-dependent manner. Neurons were treated with CHX for 0–8 h in the presence or absence of EX527 (10–50 μM). Representative western blots of 3 experiments showing the turnover of t-tau (G), ac-tau (I), or p-tau (K) in neurons with or without EX527. (H and J) The turnover of t-tau (H) or ac-tau (J) was markedly slowed by treatment of EX527. Levels of t-tau/tubulin or ac-tau/tubulin in cells harvested at time 0 were set as 1. n=3. **, P<0.01; ***, P<0.001 (two-way ANOVA, EX527-treated vs. vehicle-treated). See Figure S5 for comparison of the half-life of ac-tau vs. t-tau. Values are means ± SEM (A–B, D, G, I).

Figure 7. Tau Acetylation Is Elevated under Pathological Conditions

(A) Tau acetylation was increased by low levels of Aβ oligomers in primary cortical neurons (DIV=11). n=5 from 3 experiments. **, P=0.003 (one-way ANOVA and Tukey-Kramer posthoc test). (B) Tau acetylation was associated with familial MAPT mutations in primary neurons (DIV=13). Ac-tau/t-tau levels in neurons infected with Lenti-hTauwt were set as 1. n=9 from three experiments. *, P=0.013 (unpaired t test). (C) Representative western blots showing levels of ac-tau, t-tau, and hyperphosphorylated tau in human brains (Bm-22, superior temporal gyrus) at different Braak stages (0–5). (D) Ac-tau levels were elevated in patients with mild (Braak stages 1–2) to moderate (Braak stages 3–4) levels of tau pathology. n=8–18 cases/Braak range. *, P<0.05; **, P<0.01, one-way ANOVA Tukey-Kramer posthoc analyses. See Table-S2 for the patient information. Values are means±SEM (A, B, D).

Figure 8. Reducing Tau Acetylation Eliminates p-Tau

(A) C646 (20 μM) eliminated ac-tau and AT8-positive p-tau within 2 h in primary cortical neurons (DIV=9). Representative western blot of two experiments. (B) C646 (20 μM) eliminated p-tau. Levels of p-tau/GAPDH in non-treated cells were set as 1. n=4. ***, P<0.0001 (unpaired t test). (C) C646 (20 μM) eliminated AT8-positive p-tau in primary neurons expressing hTauP301L (DIV=12). Left, Representative western blot of two experiments. Right, Levels of p-tau/GAPDH in cells treated with control compound (C37) were set as 1. n=7. ***, P=0.0001 (unpaired t test). (D) Hypothetical model of how tau acetylation may contribute to tau-mediated neurodegeneration. Dashed lines and factors in grey indicate pathways not yet tested. Values are means±SEM (B–C).