An acetylation–phosphorylation switch that regulates tau aggregation propensity and function

The aberrant accumulation of tau protein is a pathological hallmark of a class of neurodegenerative diseases known as tauopathies, including Alzheimer's disease and related dementias. On the basis of previous observations that tau is a direct substrate of histone deacetylase 6 (HDAC6), we sought to map all HDAC6-responsive sites in tau and determine how acetylation in a site-specific manner affects tau's biophysical properties in vitro. Our findings indicate that several acetylation sites in tau are responsive to HDAC6 and that acetylation on Lys-321 (within a KCGS motif) is both essential for acetylation-mediated inhibition of tau aggregation in vitro and a molecular tactic for preventing phosphorylation on the downstream Ser-324 residue. To determine the functional consequence of this HDAC6-regulated phosphorylation event, we examined tau's ability to promote microtubule assembly and found that phosphorylation of Ser-324 interferes with the normal microtubule-stabilizing function of tau. Tau phosphorylation of Ser-324 (pSer-324) has not previously been evaluated in the context of tauopathy, and here we observed increased deposition of pSer-324–positive tau both in mouse models of tauopathy and in patients with Alzheimer's disease. These findings uncover a novel acetylation–phosphorylation switch at Lys-321/Ser-324 that coordinately regulates tau polymerization and function. Because the disease relevance of this finding is evident, additional studies are needed to examine the role of pSer-324 in tau pathobiology and to determine whether therapeutically modulating this acetylation–phosphorylation switch affects disease progression in vivo.

S1. Comparison of tau acetylation across proteomic studies.
AC sites on tau p300 (Min et al., 2010) CBP (Cohen et al., 2011) p300 current study endogenous mouse tau (Morris et al., 2015) K148 Table S1. Comparison of tau acetylation across proteomic studies. The results of the current study are compared with two previous reports which similarly utilized mass spectrometry to identify acetylated residues in tau following in vitro acetylation with either p300 [2] or CBP [1] acetyltransferase enzymes. Given that the data in the current report was collected using two different fragmentation methods, including HCD (higher energy collisional dissociation) and CID (collision-induced dissociation), modifications that were detected with only one method or were also observed in the negative reaction (no p300 enzyme) are further documented. The in vitro modification of human tau is then compared with in vivo lysine modifications observed on endogenous mouse tau in nontransgenic and APP transgenic mice [3], with residues numbered according to the longest human tau isoform. a Modification in grey font denotes changes only identified in APP transgenic mice Preparation of acetylated/deacetylated tau reactions for mass spectrometry. Recombinant tau and acetyl CoA were incubated in the absence (reaction 1) or presence (reactions 2-4) of active p300. Following the acetylation reaction, recombinant HDAC6 was added to reactions 3-4, and the HDAC6 inhibitor ACY-738 (10M) was also included in reaction 4 to inhibit deacetylation. Immunoblotting was utilized to confirm p300 induced robust tau acetylation (reaction 2), HDAC6 stimulated deacetylation of tau (reaction 3), and the addition of ACY-738 inhibited HDAC6-mediated tau deacetylation (reaction 4). Samples were then separated by SDS-PAGE, visualized by silver stain, and the bands corresponding to tau excised for analysis by mass spectrometry.

Fig. S2
Pseudoacetylation of K321 inhibits tau aggregation. a Pelleting analysis was used to compare dextran sulfate-induced tau aggregation of WT and K321Q/R mutant tau proteins. Following pelleting of insoluble tau, the pelleted fraction was separated by SDS-PAGE and detected by Coomassie blue stain. b Quantitation of the pelleting reaction revealed a significant decrease in insoluble K321Q relative to WT, while K321R exhibited an enhanced ability to aggregate and shift to the insoluble fraction (F=35.2, p<0.0005). **p<0.01 *p<0.05

Fig. S3
pS324 antibody is specific for phosphorylated S324. Recombinant WT or S324A tau was phosphorylated in vitro with PKA, and the reactions subsequently evaluated by immunoblotting. The lack of pS324-positivity in the phosphorylated S324A sample confirms that pS324 antibody is specific for phosphorylated S324, while the presence of 12E8 (pS262/356) confirms both WT and S324A mutant tau proteins were phosphorylated by PKA in vitro. Total tau levels were detected with E1.

Fig. S4
Phosphorylation of S324 is detected in rTg4510 model of tauopathy. The appearance of different phospho-tau epitopes with aging/disease progression was evaluated by immunoblotting brain lysates of rTg4510 mice at 2, 6 and 10 months of age. While 12E8 (pS262/356) and PHF1 (pS396/404) positive tau species were already present at 2 months, pS324 is absent at 2 months of age, but is detected by 6 and 10 months.

Fig. S5
Phosphorylation of KXGS motifs in AD brain. The full blots showing pS324, 12E8 and E1 immunoreactivity (Fig.5a) along with molecular weight markers is provided. HDAC6 inhibition reduces pS324 in primary neuronal cultures. The full blots with molecular weight markers and experimental replicates for the immunoblots and quantitation presented in Fig.6 is provided.