FLEXITau: Quantifying Post-translational Modifications of Tau Protein in Vitro and in Human Disease

Abstract

Tauopathies, including Alzheimer's disease (AD), are associated with the aggregation of modified microtubule associated protein tau. This pathological state of tau is often referred to as "hyperphosphorylated". Due to limitations in technology, an accurate quantitative description of this state is lacking. Here, a mass spectrometry-based assay, FLEXITau, is presented to measure phosphorylation stoichiometry and provide an unbiased quantitative view of the tau post-translational modification (PTM) landscape. The power of this assay is demonstrated by measuring the state of hyperphosphorylation from tau in a cellular model for AD pathology, mapping, and calculating site occupancies for over 20 phosphorylations. We further employ FLEXITau to define the tau PTM landscape present in AD post-mortem brain. As shown in this study, the application of this assay provides mechanistic understanding of tau pathology that could lead to novel therapeutics, and we envision its further use in prognostic and diagnostic approaches for tauopathies.

Figure 1

FLEXITau experimental workflow. (A) In a typical FLEXITau experiment the heavy tau standard is generated in the presence of heavily labeled amino acids and added to an unlabeled endogenous sample in a ratio of approximately 1:1. After enzymatic digest and LC-MS analysis all unmodified tau peptides will be observed as pairs, featuring the light and heavy isotopologues. For modified peptides, the modification causes a mass shift, reducing the amount of detectable unmodified peptide and causing a deviation of the mixing ratio. The extent of modification on that peptide can be inferred by the amount of “missing” unmodified species. (B) Plotting of peptide L/H ratios sorted from protein N- to C-terminus allows for a global visualization of modified peptides and protein regions. Blue, heavy tau; dark orange, light tau; P, phosphorylation.

Figure 2

FLEXITau sequence coverage and detection limit. (A) A dilution series of tau was performed and peptide intensities were measured using the developed FLEXITau SRM assay. log 2 peptide abundances (mean value of triplicate measurements) are shown as a heat map for the quantified peptides, sorted from N- to C-terminus. (B) Achieved sequence coverage of tau relative to injection amount. Minimal concentration for maximum sequence coverage is indicated by dotted line. (C) Representative curves of eight peptides (mean value of triplicate measurements).

Figure 3

Quantification of “hyperphosphorylated” tau by FLEXITau (A) Sf9-tau sample preparation. Sf9 insect cells were transfected with recombinant baculovirus encoding for human 2N4R wild type tau. Purified Sf9-tau (P-tau) was treated with AP to generate dephosphorylated Sf9-tau (deP-tau). Phosphatase inhibition by OA prior to cell harvest and tau purification gives rise to hyperphosphorylated Sf9 (PP-tau). As control, unmodified tau was expressed in E. coli. (B) Three independent preparations of Sf9-tau were subjected to FLEXITau SRM analysis (three SRM measurements each). Technical reproducibility (top panel) and biological reproducibility (bottom panel) of L/H ratio was calculated as %CV. Data are represented in a boxplot (5–95% whiskers; mean indicated by +). (C) Peak areas were extracted from the same data set and the %CV calculated. Data are represented as boxplot (5–95% whiskers; mean indicated by +). (D) FLEXITau data are shown for each tau species (average of three independent Sf9-tau preparations; bar shows relative error (%CV)). Peptides sorted from N- to C-terminus are projected onto a schematic of 2N4R tau protein, respective to their amino acid location. Exons prone to splicing (exons 2, 3, and 10) are depicted in dark brown. Peptides not modified in P-tau and PP-tau are shown in green; peptides significantly different from ctrl-tau in orange, and peptides different in PP-tau only in red (student t-test, p < 0.05). (E) Phosphorylated sites identified by LC-MS/MS shotgun analysis were mapped onto tau peptides that are color coded according to their modification extent as quantified by the SRM analysis. For the full list of modified peptides see SI Table S2.

Figure 4

Calculation of site occupancies for singly and multiply modified peptides. (A) For peptides containing a single modification site, the site occupancy equals the peptide modification extent calculated by the FLEXITau assay (see Table 1). For multiply modified peptides, we designed a combinatorial approach to stepwise calculate individual site occupancies by using quantitative information from overlapping peptides: (i) the proline-rich region; eqs 1–9, with eq 1 starting from N-terminus and eq 5 starting from the C-terminus; (ii) C-terminal tail, eqs 10 and 11. (B) Phosphorylation extents for all identified sites are shown in percent (average ± std dev of three biological replicates, three measurements each). Sites are sorted by the amino acid location in the tau sequence N- to C-terminus. Commonly used diagnostic AD antibodies are shown in blue, with binding epitopes underlined; dotted line indicating controversy regarding antibody epitope. The gray boxes indicate the highly modified regions i and ii shown in A. P, phosphate groups.

Figure 5

Applicability of FLEXITau to human disease. (A) Sarkosyl insoluble tau from post-mortem brains of three different AD patients was subjected to the FLEXITau workflow (triplicate SRM measurements). The modification extent of each peptide was compared to Sf9-tau (see also Figure 3D) and in-scale mapped to a schematic tau protein. The color code depicts the respective modification extent (green, no modification; red, 100% modification). (B) Peptides were sorted using hierarchical clustering of the FLEXITau data (Euclidean distance, Ward's criteria). PTMs reported on AD-tau in the literature are listed next to the corresponding peptide. If no prefix is present, the amino acid site refers to phosphorylation: ub, ubiquitination, ac, acetylation, bold, modifications detected in Sf9-tau. References: (a) Cripps et al., (b) Hanger et al., (c) Cohen et al., (d) Martin et al., (e) Thomas et al., (f) Noble et al., (g) Grinberg et al., and (h) Dammer et al.