Discovery and characterization of stable and toxic Tau/phospholipid oligomeric complexes

Abstract

The microtubule-associated protein Tau plays a central role in the pathogenesis of Alzheimer's disease. Although Tau interaction with membranes is thought to affect some of its physiological functions and its aggregation properties, the sequence determinants and the structural and functional consequences of such interactions remain poorly understood. Here, we report that the interaction of Tau with vesicles results in the formation of highly stable protein/phospholipid complexes. These complexes are toxic to primary hippocampal cultures and are detected by MC-1, an antibody recognizing pathological Tau conformations. The core of these complexes is comprised of the PHF6* and PHF6 hexapeptide motifs, the latter in a β-strand conformation. Studies using Tau-derived peptides enabled the design of mutants that disrupt Tau interactions with phospholipids without interfering with its ability to form fibrils, thus providing powerful tools for uncoupling these processes and investigating the role of membrane interactions in regulating Tau function, aggregation and toxicity.

Fig. 1

Tau and K18 bind to and disrupt vesicles formed with negatively charged phospholipids, resulting in the formation of globular Tau/K18-phospholipid complexes. a Schematic depiction of the Tau sequence. The longest Tau isoform consists of an N-terminal projection domain containing two negatively charged inserts, N1 and N2 (orange), and a proline-rich region (green), and a C-terminal microtubule-binding domain with four microtubule-binding repeats, R1−R4 (blue). R2 and R3 contain the hexapeptides PHF6* and PHF6 (purple), respectively, which are necessary to drive Tau fibrilization. b Left: schematic representation of the vesicle sedimentation assay set-up; fraction 1: sample containing no sucrose, fraction 2: interface between sample and 10% sucrose solution, fraction 3: 10% sucrose solution, fraction 4: interface between 10 and 60% sucrose solutions. Right: SDS-PAGE analysis of the different fractions for different samples. c Negative-stain EM images of BPS vesicles alone (left) and mixed with Tau (middle) and K18 (right) at a molar protein:phospholipid ratio of 1:20 immediately after mixing (top panels) and after 24 h of incubation (bottom panels). The scale bars are 100 nm

Fig. 2

Characterization of the Tau and K18 protein/phospholipid complexes. a Chromatograms of BPS vesicles alone (blue) and after incubation for 24 h with K18 (black) and Tau (red) at a molar protein:phospholipid ratio of 1:20. The samples were run on a Superose 6 column at a flow rate of 0.5 ml min⁻¹. Vesicles eluted in the void volume (15 min), globular particles formed by Tau and K18 eluted between 20 and 25 min and between 27 and 32 min, respectively, and monomeric Tau and K18 eluted at 30 min and 36 min, respectively. b Negative-stain EM images of the void volume, the globular particles, and monomer fractions of Tau (top) and K18 (bottom). The scale bars are 100 nm. c SDS-PAGE gel of the void volume (V) and the peaks representing the protein/phospholipid complexes (C) and the monomers (M) obtained with Tau (top) and K18 (bottom). The asterisks mark bands of monomeric proteins and the dashed red rectangles indicate oligomeric species. The phospholipids, containing 1% NBD-labeled phospholipids, which were imaged separately with a Typhoon trio fluorescent scanner, are shown on the bottom (labeled fPS)

Fig. 3

Structural characterization of Tau and K18 protein/phospholipid complexes. a NativePAGE analysis of protein/phospholipid complexes formed by Tau (left) and K18 (right) after 24 h of incubation with BPS vesicles at a molar protein:phospholipid ratio of 1:20. The monomer bands are indicated by asterisks, and the presence of higher molecular weight bands, indicated by arrows, demonstrates that the protein/phospholipid complexes contain oligomerized proteins. b Averages obtained from the classification of ~4850 particles of Tau with a radius ranging from 100 to 110 Å and ~4000 particles of K18 with a radius ranging from 81 to 97 Å into 50 classes. c CD spectra of Tau (left) and K18 (right) alone (blue) and their phospholipid complexes (red) formed by incubation with BPS vesicles at a molar protein:phospholipid ratio of 1:20. The formation of protein/phospholipid complexes is accompanied by a slight increase in secondary structure content. d Dot blots (0.25 µg) of Tau monomer or Tau incubated for 4 days with BPS vesicles at 37 °C. Tau was detected using the MC-1 antibody (dilution 1:500), which recognizes pathological Tau conformations, and a homemade rabbit polyclonal anti-Tau antibody (dilution 1:5000) to assess the total amount of protein. e Determination of the molecular weight of protein/phospholipid complexes formed by Tau and K18 by multi-angle light scattering. Although the number of subunits in the complex remains constant (5 for Tau and 7–8 for K18), the number of phospholipid decreases over time (from ~190 to ~150 for Tau and from ~160 to ~140 for K18). The final molecular weight of the protein/phospholipid complexes formed by Tau and K18 are ~335 kDa and 230 kDa, respectively. f Residue type assignments of [U-13C/15N]-labeled K18/phospholipid complexes using 2D DARR solid-state NMR spectra. Mixing times for DARR are 20 ms (blue) and 150 ms (red). Both spectra were recorded at 17.6 T, 278 K, and 13 kHz magic-angle spinning

Fig. 4

Tau isoforms lacking the R2 domain are deficient in the formation of protein/phospholipid complexes. a Human Tau naturally occurs in six isoforms, depending on the alternative splicing of exons E2, E3, and E10, giving rise to the variants containing either 0, 1, or 2 N-terminal inserts (0N, 1N, and 2N), and 3 or 4 microtubule-binding repeats (3R and 4 R). b NativePAGE analysis of protein incubated alone or with BPS vesicles at a molar protein:phospholipid ratio of 1:20. The Tau 0N4R isoform was able to form protein/phospholipid complexes as efficiently as the WT protein, while Tau 0N3R and K19 formed significantly less complexes. The monomer bands, are indicated by asterisks, and the presence of higher molecular weight bands are indicated by arrows. c Co-sedimentation assay of BPS vesicles mixed with the three Tau isoforms (from top to bottom: Tau 2N4R, Tau 0N4R, and Tau 0N3R) at a molar protein:phospholipid ratio of 1:100. When incubated alone (left), all proteins remained in the supernatant (sup.), but in the presence of vesicles (right), the proteins partially co-sedimented with the vesicles to the pellet fraction (pel.), indicative of membrane binding. Tau 2N4R and Tau 0N4R co-sedimented more than Tau 0N3R. d EM analysis of Tau 0N4R (left panel), Tau 0N3R (middle panel) and K19 (right panel) mixed with BPS vesicles at a molar protein:phospholipid ratio of 1:20 showed that Tau 0N4R can form protein/phospholipid complexes, while Tau 0N3R formed some large round and bean-like structures, and K19 formed few complexes with many vesicles remaining. The scale bars are 100 nm

Fig. 5

Identification of residues that form the K18/phospholipid oligomer core. a Valine and isoleucine spin system assignments showing the presence of two Val residues and one Ile residue in a β-strand conformation (signals shifted down and right reflect decreased Cα and increased Cβ shifts characteristic of β-strand). b Proton-nitrogen HSQC spectrum of a K18/phospholipid complex preparation in solution (red), compared with a matching spectrum of the monomeric protein (black). Resonance assignments are indicated, with red labels for the PHF6 and PHF6* regions. c Secondary carbon chemical shifts (Cα-Cα_rc + Cβ-Cβ_rc) by amino acid type for signals observed in DARR spectra of K18/phospholipid complexes. Lys-2 is marked with an asterisk to indicate that this tentatively assigned lysine residue exhibits chemical shifts consistent with a pre-proline residue, indicating that it corresponds to residue Lys-311. d–f Regions from the proton-nitrogen correlation HSQC spectra of complexes suspended in solution (red), compared with matching spectra of the monomeric protein (black). The signals from the PHF6 region (Tyr-310 and Leu-315), from the PHF6* region (Gln-276), and from the ~9 residues subsequent to each PHF motif (Val-287 and Val-318) are highly attenuated, indicating their immobilization in the oligomer core. Signals from other regions of K18 are much less affected, indicating that they remain highly flexible outside the oligomer core. Signals from Ile-297 and Ile-328 also show some attenuation, possibly resulting from their close proximity to regions within the core, which likely restricts their mobility. g Schematic of the K18 sequence and its location within full-length Tau, highlighting in red the residues that appear to form the oligomer core. The hexapeptide motifs PHF6 and PHF6* are underlined and in bold type. h Ratios of intensities for well resolved signals in HSQC spectra of protein/phospholipid complexes versus monomers, showing that the PHF6 region and subsequent ~9 residues, are most highly attenuated, but that the corresponding region in repeat 2, including the PHF6* motif and subsequent ~9 residues are also significantly attenuated compared to the remainder of the protein

Fig. 6

Neuronal internalization of Tau and K18 monomers, fibrils and protein/phospholipid complexes. Hippocampal primary neurons were treated with PBS, 1 µM Tau or Oregon Green-labeled K18 (K18-OG) protein/phospholipid complexes for 1 day and processed for immunocytochemistry. From left to right: overview image, DAPI staining to show the nucleus (blue), K18-OG or staining with anti-Tau antibody Tau13 (green), staining with anti-MAP2 antibody identifying neurons (red), and orthogonal projections. In neurons treated with protein/phospholipid complexes containing K18-OG, K18 localizes to the plasma membrane and is internalized by primary neurons after 1 day of treatment, as evidenced by the co-staining with MAP2. In neurons treated with monomeric K18-OG, K18 is internalized and diffuse in the cytoplasmic compartment and in small punctae after 1 day of treatment. In neurons treated with K18-OG fibrils, K18 localizes to the membrane in the form of large punctae that are internalized by primary neurons after 1 day of treatment. In neurons treated with Tau/phospholipid complexes, Tau localizes to the plasma membrane and is internalized after 1 day of treatment. In neurons treated with monomeric Tau, Tau is internalized and present diffuse in the cytoplasmic compartment and in small punctae after 1 day of treatment. In neurons treated with sonicated Tau fibrils, Tau localizes to the plasma membrane and is not internalized at 1 day post-treatment. The scale bars are 20 µm (low magnification) and 5 µm (high magnification)

Fig. 7

Design and characterization of the membrane binding-deficient K18 and Tau mutants. a Helical wheel and β-strands depictions to illustrate the rational for the design of the K18 and Tau mutants. Mutations are in purple, hydrophilic residues in blue, hydrophobic residues in black, positively charged residues in red, cysteine in green, and glycine in white. The hydrophobic and hydrophilic faces of the helices are depicted in light gray and blue, respectively, and the putative position of the negatively charged phospholipid head groups are indicated by dashed lines. Mutation of a hydrophobic Val residue in the hydrophobic part of an amphipathic motif to a negatively charged Glu should interfere with the capacity of the segment to embed itself into the hydrophobic core of a membrane. In the case of Tau, the modification of positively charged Lys-311 to an uncharged Ala should affect its capacity to interact with negatively charged membranes through electrostatic interactions. b Co-sedimentation assay of BPS vesicles with WT and MBD-K18 (top) and Tau (bottom). When incubated alone (left), all proteins remain in the supernatant (sup.), but in the presence of BPS vesicles (right), the proteins partially co-sediment with the vesicles to the pellet (pel.), indicative of membrane binding. Compared with the WT proteins, more MBD-K18 and MBD-Tau proteins remain in the supernatant, indicating that they partially lost the ability to bind membranes. c Top: size-exclusion chromatograms of WT K18 (solid lines) and MBD-K18 (dashed lines) alone (blue lines) or with BPS vesicles (red lines) show that only WT K18 forms protein/phospholipid complexes. Bottom: size-exclusion chromatograms of WT Tau (blue line) and MBD-Tau (red line) show that WT Tau is efficient in forming protein/phospholipid complexes (blue, large peak) while the capacity of MBD-Tau to form such complexes is compromised (red, small peak). d EM analysis of WT and MBD-K18 (top) and Tau (bottom) shows that the mutants cannot form protein/phospholipid complexes in the presence of BPS vesicles (second-to-left panels) but retain their ability to form fibrils (right-most panels). The scale bars are 100 nm

Fig. 8

Tau and K18 protein/phospholipid complexes, fibrils, and monomers are toxic to hippocampal neurons and exert their toxicity through different mechanisms. Hippocampal mouse primary neurons were treated with PBS, BPS vesicles (controls), 3 µM Tau or K18 monomers, protein/phospholipid complexes or sonicated fibrils and fixed at 1 or 3 days post treatment. a Apoptotic cells were stained using the TUNEL assay, the neuronal population was stained using NeuN, and the nucleus was stained using DAPI. The percentage of apoptotic neurons [(TUNEL + and NeuN + cells)/NeuN + cells] was quantified as described in the methods. No statistical significance was observed at 1 day, whereas all species were significantly toxic at 3 days. Bars are mean ± S.D. b Tau and K18 monomers, fibrils, and protein/phospholipid complexes exert different effects on primary hippocampal neuron morphology. At 3 days, the PBS-treated and BPS-treated neurons present a healthy neuronal morphology with round cell bodies and extended neurites (left panels). Neurons treated with monomeric Tau and K18 show many condensed neurons (second left panels, arrows). Neurons treated with K18 or Tau fibrils clumped together, linked by bundles of neurites (second right panels, arrows). Tau/phospholipid complexes presented both condensed and clumped cell bodies, with bundled neurites (right bottom panel, arrows). K18/phospholipid complexes induced condensation of cell bodies and shrinkage of the neurites (right top panel, arrows). c Apoptotic cell death was confirmed using the CaspaTag fluorescein caspase 3 activity kit (Axxora) as described in the methods section. All species exhibited significant toxicity at 3 days post treatment. Bars are mean ± S.E.M. d The level of apoptosis induced by MBD-Tau and MBD-K18 species was assessed using the caspase 3 activity kit as described in the methods. The level of toxicity was higher than that of the BPS-treated neurons, but toxicity was only significant in the case of the MBD-K18 fibrils. Bars are mean ± S.E.M. (*p < 0.05, **p < 0.01, ***p < 0.001)

Fig. 9

Schematic depiction of our working model that summarizes our findings for the interaction of Tau with negatively charged phospholipids. Tau forms fibrils readily in the presence of heparin, and Tau monomers interact with vesicles containing negatively charged phospholipids, potentially though the formation of short amphipathic α-helical (green) and β-sheet (purple) motifs^,,. Subsequently, phospholipid molecules are segregated into highly stable protein/phospholipid complexes, possibly through a tweezers mechanism (left). The core of these protein/phospholipid complexes consists of PHF6 and PHF6*, and regions C-terminal to these two motifs, whereby only the PHF6 is folded into a β-sheet, while the rest of the core remains unfolded. Based on these findings, mutants were designed that retain the ability to from fibrils in the presence of heparin, but no longer interact with phospholipids (right) to form protein/phospholipid complexes, thus providing powerful tools to investigate how membrane binding affects the normal and pathological functions of Tau