A Dual Pathogenic Mechanism Links Tau Acetylation to Sporadic Tauopathy

Abstract

Tau acetylation has recently emerged as a dominant post-translational modification (PTM) in Alzheimer's disease (AD) and related tauopathies. Mass spectrometry studies indicate that tau acetylation sites cluster within the microtubule (MT)-binding region (MTBR), suggesting acetylation could regulate both normal and pathological tau functions. Here, we combined biochemical and cell-based approaches to uncover a dual pathogenic mechanism mediated by tau acetylation. We show that acetylation specifically at residues K280/K281 impairs tau-mediated MT stabilization, and enhances the formation of fibrillar tau aggregates, highlighting both loss and gain of tau function. Full-length acetylation-mimic tau showed increased propensity to undergo seed-dependent aggregation, revealing a potential role for tau acetylation in the propagation of tau pathology. We also demonstrate that methylene blue, a reported tau aggregation inhibitor, modulates tau acetylation, a novel mechanism of action for this class of compounds. Our study identifies a potential "two-hit" mechanism in which tau acetylation disengages tau from MTs and also promotes tau aggregation. Thus, therapeutic approaches to limit tau K280/K281 acetylation could simultaneously restore MT stability and ameliorate tau pathology in AD and related tauopathies.

Figure 1. Tau acetylation modulates tau phosphorylation at specific epitopes.

(a–c) Immunoblot analysis and quantification using the indicated tau antibodies for (a) cells expressing WT tau-T40 and either an active form of CBP or a catalytically inactive form of CBP (CBP-LD), (b) cells co- expressing tau-T40 (WT, ΔKK, 4KQ, or 4KR) and tau modifying enzymes, and (c) cells expressing tau-T40 (WT, ΔKK, 2KQ, 4KQ, or 4KR) and treated with control (DMSO) or okadaic acid (OA), where indicated. Solid black arrows highlight the ~75 kDa phospho-tau species that is reduced upon acetylation. Statistical significance was assessed using a student t-test (****p < 0.0001). Cropped images from full size immunoblots are provided in panels a–c. Full-length immunoblots are presented in Supplementary Fig. S1. (d) Immunofluorescence microscopy of primary cortical neurons expressing WT-tau-GFP (left) or 4KQ-tau-GFP (right) detecting phosphorylated tau (AT8). White arrows identify transfected neurons. Scale bar, 50 µm.

Figure 2. Tau acetylation-mimics impair tau-mediated MT assembly.

(a) Peptide spectra depicting doubly acetylated K280 and K281 (²⁷⁵VQIINKKLDLSNVQSK²⁹⁰) from CBP-acetylated tau proteins. Ion scores are provided in Supplementary Fig. S2. (b,c) Autoradiography and Coomassie blue staining of (b) WT 4R-tau-K18 and 3 R-tau-K19 or (c) WT tau-K18 and tau-K18 ΔKK (containing deletion of residues K280 and K281). Cropped images from full size gels are provided in panels (b,c). Full-length gels are presented in Supplementary Fig. S3. (d,e) Light scattering assay in the presence of tau-K18 proteins detecting tubulin polymerization, as determined by absorbance readings at 350 nm (No Tau, X (N = 7); WT, ♦ (N = 9); K280Q, ● (N = 8); 2KQ, ▲ (N = 4); K280R, ◯ (N = 7); 2KR, Δ (N = 6)). Error bars indicate s.d. of the mean. Statistical significance was assessed using a student t-test as follows: WT vs. K280Q (***p < 0.0001); WT vs. K280R (ns, p = 0.36); K280Q vs. K280R (***p < 0.0001); WT vs. 2KR (**p < 0.0072); WT vs. 2KQ (***p < 0.0001); 2KQ vs. 2KR (***p < 0.0001).

Figure 3. Tau K280 acetylation-mimic enhances tau aggregation in vitro.

(a) Coomassie blue staining of monomeric supernatant (S) and fibrillar pellet (P) fractions of tau-K18 fibril reactions (0 to 8 h). (b) Thioflavin T (ThT) fluorescence of tau-K18 fibril reactions at the indicated time points from 0 to 8 h. Error bars indicate s.d of the mean. (c) Coomassie blue staining of monomeric supernatant (S) and fibrillar pellet (P) fractions of full-length tau-T40 fibril reactions (0 to 3 d). (d) ThT fluorescence of tau-T40 fibril reactions at the indicated time points from 0 to 3 d. Error bars indicate s.d of the mean derived from N = 3 independent experiments. Cropped images from full size gels are provided in panels a and c. Full-length gels are presented in Supplementary Fig. S4. Error bars indicate s.d. of the mean. Statistical significance was assessed using a student t-test (*p < 0.05; **p < 0.01; ***p < 0.001).

Figure 4. Tau K280 acetylation-mimic fibrils show similar ultrastructural properties to P301L mutant tau.

Negative staining electron microscopy (EM) of tau-K18 protein fibril reactions incubated for 2 h (top row) and 4 h (bottom row). Scale bar, 500 nm.

Figure 5. Tau K280 acetylation-mimic accelerates the formation of β-structure.

(a–d) CD spectra of tau-K18 fibril reactions at the indicated time points from 0 to 1 h (0 h, black line; 1 h, grey line) were recorded from 190 to 250 nm. 0 h time points indicate monomeric soluble tau (random coil) and 1 h time point indicates fibrillar tau derived from the pellet fraction. Error bars indicate standard deviation (s.d.) of the mean derived from N = 3 independent experiments.

Figure 6. Tau acetylation-mimics enhance seed-dependent tau aggregation in cells.

(a) Total and phospho-tau immunoblotting of full-length tau-T40 expressing QBI-293 cells (WT, P301L, 2KQ, 2KR) seeded with K18-PL fibrils or, as a negative control, no fibril treatment (no fib). (b) Quantification of immunoblots in (a) was performed by protein band densitometry. Cropped images from full size immunoblots are provided in panel a. Full-length immunoblots and additional quantification are presented in Supplementary Fig. S5. Statistical significance was assessed using a student t-test (*p < 0.05; ***p < 0.001).

Figure 7. Methylene blue (MB) suppresses the accumulation of tau K280 acetylation-mimic.

(a) Coomassie blue staining of monomeric supernatant (S) and fibrillar pellet (P) fractions of tau-K18 WT and K280Q fibril reactions in the absence of presence of MB. (b) Immunoblot analysis with indicated antibodies of cells expressing K18-WT or K18-K280Q treated with 3MA, Ars, and/or MB, where indicated. Hsc70 was included as a loading control. (c) Quantification of immunoblots in (b). Error bars indicate s.d. of the mean. Statistical significance was assessed using a student t-test (*p < 0.05; **p < 0.01; ***p < 0.001). Cropped images from full size gels and immunblots are provided in panels a and b. Full-length gels and immunoblots are presented in Supplementary Fig. S6.