Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat
Sign Up
Cited by 106Alzheimer's disease research and treatmentsAmyotrophic Lateral Sclerosis ResearchMicrotubule and mitosis dynamics

The DOI matches its Crossref record; 94 of the 94 references it cites were found in a registry or an index; none of the 93 sources its claims cite is checked for support yet.

Molecular crowding and RNA synergize to promote phase separation, microtubule interaction, and seeding of Tau condensates

Open access
Hochmair, Janine, Exner, Christian, Franck, Maximilian, Dominguez‐Baquero, Alvaro, Diez, Lisa, Brognaro, Hévila, Kraushar, Matthew L, Mielke, Thorsten, Radbruch, Helena, Kaniyappan, Senthilvelrajan, Falke, Sven, Mandelkow, Eckhard, Betzel, Christian, Wegmann, Susanne·2022·The EMBO Journal
DOI 10.15252/embj.2021108882PMC PMC9156969
Open this paper in CoChat — free accountOpen PDF

Abstract

Biomolecular condensation of the neuronal microtubule‐associated protein Tau (MAPT) can be induced by coacervation with polyanions like RNA, or by molecular crowding. Tau condensates have been linked to both functional microtubule binding and pathological aggregation in neurodegenerative diseases. We find that molecular crowding and coacervation with RNA, two conditions likely coexisting in the cytosol, synergize to enable Tau condensation at physiological buffer conditions and to produce condensates with a strong affinity to charged surfaces. During condensate‐mediated microtubule polymerization, their synergy enhances bundling and spatial arrangement of microtubules. We further show that different Tau condensates efficiently induce pathological Tau aggregates in cells, including accumulations at the nuclear envelope that correlate with nucleocytoplasmic transport deficits. Fluorescent lifetime imaging reveals different molecular packing densities of Tau in cellular accumulations and a condensate‐like density for nuclear‐envelope Tau. These findings suggest that a complex interplay between interaction partners, post‐translational modifications, and molecular crowding regulates the formation and function of Tau condensates. Conditions leading to prolonged existence of Tau condensates may induce the formation of seeding‐competent Tau and lead to distinct cellular Tau accumulations. Condensation of the microtubule‐associated protein Tau is induced by the presence of polyanions like RNA or by molecular crowding. Here we show that these conditions have synergistic effects on Tau condensation and Tau‐dependent microtubule bundling in vitro, and on the seeding of Tau aggregates in cells. Conditions that induce liquid‐liquid phase separation have distinct effects on Tau condensation and subcellular localization, including formation of Tau‐positive inclusions along the nuclear envelope that correlate with nucleocytoplasmic transport deficits.

Overview

CoChat Summary

TL;DR

Molecular crowding and RNA synergize to enable Tau condensation at physiological ion conditions and produce condensates with strong affinity to charged surfaces. Aged Tau condensates, including non-percolating Tau:RNA coacervates, seed pathological Tau accumulations in cells, including nuclear-envelope Tau linked to nucleocytoplasmic transport deficits.

Key findings

  • Crowding plus RNA stabilized Tau condensates against salt: Tau:RNA coacervates disappeared at 25 mM NaCl, Tau:PEG condensates destabilized at 100 mM NaCl, but Tau:PEG:RNA condensates did not change between 5 and 150 mM NaCl and still formed at 300 mM NaCl. In axoplasm-like X/2 buffer, Tau:PEG formed condensates and Tau:PEG:RNA formed more robustly, while Tau:RNA coacervation was inhibited. Growth kinetics gave coarsening exponents b = 0.44 for Tau:PEG:RNA, b = 0.15 for Tau:RNA, and b = 0.18 for Tau:PEG.
  • Phosphorylated Tau failed to coacervate with RNA alone but condensed with PEG; in p-Tau:PEG:RNA systems RNA co-partitioned into condensates and condensates formed at elevated NaCl. PKA-phosphorylated Tau showed the highest in-solution LLPS and surface wetting in Tau:PEG:RNA, while pT205 and pS404 had little effect or inhibited condensation and pS262/pS356/pS324 appeared to favor LLPS.
  • Tubulin co-partitioned into Tau:PEG, Tau:RNA, and Tau:PEG:RNA condensates, with RNA and tubulin competing for co-condensation. Tau:PEG:RNA produced thick MT bundles arranged around persisting condensates, whereas Tau:PEG produced thin, partly curved bundles; nocodazole stability was higher for Tau:PEG than Tau:PEG:RNA MTs.
  • 24 h-old Tau:PEG, Tau:RNA, and Tau:PEG:RNA condensates formed SDS-stable Tau dimers/oligomers and increased TauΔK280 aggregation in Thioflavin-T assays. Tau:RNA coacervates induced Tau aggregation in HEK sensor cells more efficiently than PEG-induced condensates and similar to Tau:PEG:RNA; RNA-containing condensates induced nuclear-envelope Tau with the highest frequency (26%), while heparin-induced aggregates mostly seeded cytoplasmic accumulations (74% CYT, 18% NE, 8% NUC). NE Tau had faster FRAP (t1/2 ≈ 2.5 s) than CYT Tau (t1/2 ≈ 7 s), and FLIM distinguished soluble Tau (LT-1 ~2.3 ns), NE Tau (LT-2 ~2.2 ns), CYT/NUC Tau (LT-3 ~2.1 ns), and a subset of CYT Tau (LT-4 ~2.0 ns).

Methods

  • In vitro LLPS used recombinant human full-length Tau (2N4R, 5 µM), p-Tau from Sf9 cells, TauΔK280, and in vitro phosphorylated Tau variants; condensation was induced with polyA, tRNA, heparin, PEG8000, or combinations. Main techniques included time-resolved dynamic light scattering, light/confocal microscopy, turbidity, FRAP, FLIM, sedimentation assays, SDS–PAGE, semi-native PAGE, dot blots, and Thioflavin-T aggregation assays.
  • Cellular and tissue work used HEK293 TauRDP301S-CFP/YFP sensor cells treated with 24 h-old Tau condensates or AD brain lysate, primary mouse hippocampal neurons expressing TauRDP301S-CFP/YFP, and human AD/control brain sections. Analyses included confocal imaging, FRAP, FLIM, immunofluorescence, and statistics by Student’s t-test or one-way ANOVA with Tukey/Holmes–Sidak correction.

Limitations

  • The paper states that effects of individual or grouped Tau post-translational modifications need further investigation due to their complexity, and that preferential induction of certain Tau accumulation types by different seeds is not conclusive at this point. It also notes that NE Tau fusion was not observed, possibly due to time/spatial resolution limits or progressed percolation, and that in vitro versus cellular differences (crowding, ion composition, pH, other biomolecules; TauRD versus full-length Tau) may affect FLIM comparisons.
  • Evident from the text: many mechanistic conclusions rely on in vitro reconstitution with purified proteins and on HEK sensor cells, with 5% PEG8000 used to induce LLPS at physiological Tau concentrations; the work does not directly demonstrate endogenous Tau condensates in neurons, and cellular seeding assays use lipofectamine and tagged TauRD constructs.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.15252/embj.2021108882

Cites

94 of 94 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Kinetics of Tau condensation and mesoscopic cluster formation Domain structure of the longest human Tau isoform (2N4R) used in this study: the positively charged C‐terminal half of Tau contains four repeats (R1‐R4) and two proline‐rich regions (P1, P2), and comprises the microtubule binding region of Tau. This region binds to polyanionic macromolecules such as RNA through electrostatic interactions. The N‐terminal ~200 amino acids (aa) project from the microtubule surface and contain two negatively charged N‐terminal inserts (N1, N2). The homogenous charge distribution (sliding window of 10 aa) along its aa sequence assigns Tau an amphiphilic character.Addition of polyA RNA, PEG (PEG8000), or both induces the de‐mixing of Tau at physiological concentrations (5 μM Tau in 25 mM HEPES, 1 mM DTT, pH 7.4). Scale bars = 10 μm.Confocal fluorescence microscopy shows co‐condensation of polyA‐Cy3 (10:1, polyA:polyA‐Cy3) with Tau (5 µM Tau with 1% Tau‐DyLight488). PEG8000‐Dylight647 is excluded from Tau condensates. Scale bar = 10 μm.Representative size distribution of molecular Tau assemblies during Tau LLPS measured by time‐resolved DLS (trDLS). At 25 μM Tau in low salt (10 mM NaCl) buffer, trDLS detects Tau monomers (radii ~10 nm) and mesoscopic Tau clusters (radii ~100–200 nm). Addition of PEG, polyA, or both after ~15 min induced the formation of Tau condensates with a radius of >103 nm, at the cost of monomers and mesoscopic clusters in the solution. Data point bubble sizes correspond to DLS amplitudes, which are proportional to the intensity of scattered light of the respective particles. Data shown are the same as in Fig 2D for 0 <t<45 min.Kinetics of Tau:PEG, Tau:RNA, and Tau:PEG:RNA condensate growth assessed through power law, R = a*tb, fitting to trDLS data starting at the time of LLPS inducer addition.Kinetics of cluster growth in low salt (25 mM HEPES, 10 mM NaCl) at 30 μM and 50 μM Tau using the same fitting model as in (E).trDLS of Tau:PEG condensation at increasing NaCl concentrations.Radius of Tau:PEG condensates at different NaCl concentrations. N = 3 experiments, 10 last data points per experiments shown, data shown as mean ± SEM.Radius of Tau clusters at different NaCl concentrations. N = 2–3 experiments, 3 last data points per experiments shown, data shown as mean ± SEM.Radius of Tau monomers at different NaCl concentrations. N = 2–3 experiments, 3 last data points per experiments shown, data shown as mean ± SEM.Model of Tau condensation: Tau monomers (grey circles) and clusters coacervate with RNA (pink) to form condensates that grow through droplet fusion (Oswald ripening). Molecular crowding agents, like PEG (blue), are excluded from the dense phase. Inset: excluded volume effects exerted by PEG force individual Tau molecules to interact, thereby trigger LLPS and stabilize Tau:RNA coacervates against the electrostatic shielding of molecular interactions inside the dense phase. As a result, Tau:PEG:RNA condensates can form at cytosol‐like ion concentrations. R0 = radius of hydration. Data information: Data in H, I, J have been compared by one‐way ANOVA with Tukey test for multiple comparison. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
Figure 1 Kinetics of Tau condensation and mesoscopic cluster formation Domain structure of the longest human Tau isoform (2N4R) used in this study: the positively charged C‐terminal half of Tau contains four repeats (R1‐R4) and two proline‐rich regions (P1, P2), and comprises the microtubule binding region of Tau. This region binds to polyanionic macromolecules such as RNA through electrostatic interactions. The N‐terminal ~200 amino acids (aa) project from the microtubule surface and contain two negatively charged N‐terminal inserts (N1, N2). The homogenous charge distribution (sliding window of 10 aa) along its aa sequence assigns Tau an amphiphilic character.Addition of polyA RNA, PEG (PEG8000), or both induces the de‐mixing of Tau at physiological concentrations (5 μM Tau in 25 mM HEPES, 1 mM DTT, pH 7.4). Scale bars = 10 μm.Confocal fluorescence microscopy shows co‐condensation of polyA‐Cy3 (10:1, polyA:polyA‐Cy3) with Tau (5 µM Tau with 1% Tau‐DyLight488). PEG8000‐Dylight647 is excluded from Tau condensates. Scale bar = 10 μm.Representative size distribution of molecular Tau assemblies during Tau LLPS measured by time‐resolved DLS (trDLS). At 25 μM Tau in low salt (10 mM NaCl) buffer, trDLS detects Tau monomers (radii ~10 nm) and mesoscopic Tau clusters (radii ~100–200 nm). Addition of PEG, polyA, or both after ~15 min induced the formation of Tau condensates with a radius of >103 nm, at the cost of monomers and mesoscopic clusters in the solution. Data point bubble sizes correspond to DLS amplitudes, which are proportional to the intensity of scattered light of the respective particles. Data shown are the same as in Fig 2D for 0 <t<45 min.Kinetics of Tau:PEG, Tau:RNA, and Tau:PEG:RNA condensate growth assessed through power law, R = a*tb, fitting to trDLS data starting at the time of LLPS inducer addition.Kinetics of cluster growth in low salt (25 mM HEPES, 10 mM NaCl) at 30 μM and 50 μM Tau using the same fitting model as in (E).trDLS of Tau:PEG condensation at increasing NaCl concentrations.Radius of Tau:PEG condensates at different NaCl concentrations. N = 3 experiments, 10 last data points per experiments shown, data shown as mean ± SEM.Radius of Tau clusters at different NaCl concentrations. N = 2–3 experiments, 3 last data points per experiments shown, data shown as mean ± SEM.Radius of Tau monomers at different NaCl concentrations. N = 2–3 experiments, 3 last data points per experiments shown, data shown as mean ± SEM.Model of Tau condensation: Tau monomers (grey circles) and clusters coacervate with RNA (pink) to form condensates that grow through droplet fusion (Oswald ripening). Molecular crowding agents, like PEG (blue), are excluded from the dense phase. Inset: excluded volume effects exerted by PEG force individual Tau molecules to interact, thereby trigger LLPS and stabilize Tau:RNA coacervates against the electrostatic shielding of molecular interactions inside the dense phase. As a result, Tau:PEG:RNA condensates can form at cytosol‐like ion concentrations. R0 = radius of hydration. Data information: Data in H, I, J have been compared by one‐way ANOVA with Tukey test for multiple comparison. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.CC BY-NC-ND · PubMed Central
Figure EV1 trDLS data of Tau coacervation Averaged experimental autocorrelation function (black circles) and data fit applying CONTIN analysis (red line) of example trDLS data. Similar data was recorded and fitted for each 20 s interval of the trDLS measurements.Distribution of hydrodynamic radii derived from autocorrelation data example in (A).Representative trDLS of Tau condensation upon addition of heparin, tRNA, and polyU RNA. Data point bubble sizes correspond to DLS amplitudes, which are proportional to the intensity of scattered light of the respective particles. Kinetics of condensate growth are assessed through power law, R = a*tb, fitting to trDLS data starting at the time of LLPS inducer addition. Data shown as mean ± SD. A representative single experiment is show.
Figure EV1 trDLS data of Tau coacervation Averaged experimental autocorrelation function (black circles) and data fit applying CONTIN analysis (red line) of example trDLS data. Similar data was recorded and fitted for each 20 s interval of the trDLS measurements.Distribution of hydrodynamic radii derived from autocorrelation data example in (A).Representative trDLS of Tau condensation upon addition of heparin, tRNA, and polyU RNA. Data point bubble sizes correspond to DLS amplitudes, which are proportional to the intensity of scattered light of the respective particles. Kinetics of condensate growth are assessed through power law, R = a*tb, fitting to trDLS data starting at the time of LLPS inducer addition. Data shown as mean ± SD. A representative single experiment is show.CC BY-NC-ND · PubMed Central
Figure 2 Ion strength dependence of Tau coacervation is overcome by crowding Microscopy of Tau:polyA, Tau:PEG, and Tau:PEG:polyA condensates (5 µM Tau, 10% PEG, 4.8 µg/ml polyA) formed at increasing NaCl concentrations and in X/2‐buffer (Song & Brady, 2013) mimicking the axoplasm ion composition. Scale bar = 10 μm.trDLS of Tau condensation in X/2‐buffer.Condensate growth kinetics in X/2 buffer assessed through power law, R = a*tb, fitting to trDLS data starting at the time of LLPS inducer addition.Representative trDLS of Tau condensation in low salt buffer, with subsequent stepwise increase of NaCl. Data information: In B and D, data point bubble sizes correspond to DLS amplitudes. Data shown for 10 mM NaCl (t = 0–45 min) are the same as shown in Fig 1D.
Figure 2 Ion strength dependence of Tau coacervation is overcome by crowding Microscopy of Tau:polyA, Tau:PEG, and Tau:PEG:polyA condensates (5 µM Tau, 10% PEG, 4.8 µg/ml polyA) formed at increasing NaCl concentrations and in X/2‐buffer (Song & Brady, 2013) mimicking the axoplasm ion composition. Scale bar = 10 μm.trDLS of Tau condensation in X/2‐buffer.Condensate growth kinetics in X/2 buffer assessed through power law, R = a*tb, fitting to trDLS data starting at the time of LLPS inducer addition.Representative trDLS of Tau condensation in low salt buffer, with subsequent stepwise increase of NaCl. Data information: In B and D, data point bubble sizes correspond to DLS amplitudes. Data shown for 10 mM NaCl (t = 0–45 min) are the same as shown in Fig 1D.CC BY-NC-ND · PubMed Central
Figure EV2 Effect of phosphate on Tau condensation trDLS of Tau:PEG:polyA, Tau:PEG, and Tau:polyA condensation, with a subsequent stepwise increase of phosphate ions. Tau:polyA condensates disappear at 25 mM phosphate ions in the buffer. Tau:PEG condensates dissolve at 75 mM, and Tau:PEG:polyA at 100 mM phosphate. Data point bubble sizes correspond to DLS amplitudes, which are proportional to the intensity of scattered light of the respective particles. Dashed vertical lines indicate times of phosphate addition.Mesoscopic cluster growth at 10 mM phosphate in the buffer, at 25 μM and 50 μM Tau. Data were derived from trDLS measurements before the addition of LLPS inducers. Kinetics of cluster growth were assessed through power law, R = a*tb, fitting to trDLS data.trDLS of p‐Tau (from insect cells) condensation in p‐Tau:PEG (left) and pTau:polyA (right) systems. Power law fit to condensate growth shows a low coarsening coefficient of b = 0.12. polyA fails to induce coacervates of pTau. Data point bubble sizes correspond to DLS amplitudes, which are proportional to the intensity of scattered light of the respective particles. Kinetics of condensate growth were assessed through power law, R = a*tb, fitting to trDLS data starting at the time of LLPS inducer addition. Data shown as mean ± SD.Microscopy of pTau:PEG and pTau:PEG:polyA condensates at increasing NaCl concentrations and in X/2 buffer (only for pTau:PEG:polyA). Scale bars = 10 μm.Coomassie staining of SDS–PAGE gel showing the phospho‐Tau variants (1 μg Tau loaded per lane).
Figure EV2 Effect of phosphate on Tau condensation trDLS of Tau:PEG:polyA, Tau:PEG, and Tau:polyA condensation, with a subsequent stepwise increase of phosphate ions. Tau:polyA condensates disappear at 25 mM phosphate ions in the buffer. Tau:PEG condensates dissolve at 75 mM, and Tau:PEG:polyA at 100 mM phosphate. Data point bubble sizes correspond to DLS amplitudes, which are proportional to the intensity of scattered light of the respective particles. Dashed vertical lines indicate times of phosphate addition.Mesoscopic cluster growth at 10 mM phosphate in the buffer, at 25 μM and 50 μM Tau. Data were derived from trDLS measurements before the addition of LLPS inducers. Kinetics of cluster growth were assessed through power law, R = a*tb, fitting to trDLS data.trDLS of p‐Tau (from insect cells) condensation in p‐Tau:PEG (left) and pTau:polyA (right) systems. Power law fit to condensate growth shows a low coarsening coefficient of b = 0.12. polyA fails to induce coacervates of pTau. Data point bubble sizes correspond to DLS amplitudes, which are proportional to the intensity of scattered light of the respective particles. Kinetics of condensate growth were assessed through power law, R = a*tb, fitting to trDLS data starting at the time of LLPS inducer addition. Data shown as mean ± SD.Microscopy of pTau:PEG and pTau:PEG:polyA condensates at increasing NaCl concentrations and in X/2 buffer (only for pTau:PEG:polyA). Scale bars = 10 μm.Coomassie staining of SDS–PAGE gel showing the phospho‐Tau variants (1 μg Tau loaded per lane).CC BY-NC-ND · PubMed Central
Figure 3 Phosphorylation modulates Tau condensation and condensate affinity to surfaces Recombinant phosphorylated Tau expressed in insect cells (p‐Tau) de‐mixes in the presence of PEG but fails to coacervate with RNA (5 μM p‐Tau, 10% PEG). In the presence of PEG, polyA‐Cy3 co‐condensates with p‐Tau (arrowheads). Scale bar = 10 μm.Tau phospho‐variants generated by in vitro phosphorylation of non‐phosphorylated Tau from E. coli with different kinases (PKA, GSK3β, or Cdk5) show differences in phosphorylation at specific phospho‐site (pT205, pS262/pS356, pS324, and pS404) tested by Western blot. Non‐phosphorylated Tau and p‐Tau from insect cells are loaded for comparison. Position of tested p‐sites is indicated in the Tau sequence cartoon. Western blot for total Tau (Tau13 antibody) confirms equal loading (0.2 μg Tau per lane) of the Tau preparations.Phos‐tag gel shows degree and heterogeneity of phosphorylation in the different Tau phospho‐variants. Intensity along the lanes are plotted as colored lines next to gel image.Condensation of Tau phospho‐variants in the Tau:PEG, Tau:RNA (polyA), and pTau:PEG:RNA systems was determined by turbidity measurements for condensates in solution (absorption at 600 m) and by microscopy for surface attached condensates (surface coverage; top left). Turbidity measurements are shown on the left y‐axis (blue circles), and surface coverage in % on right y‐axis (black data points). Scale bars = 20 μm. N = 4 experiments, data points represent the average of 5 data points from turbidity measurements taken at 20 ± 5 min after LLPS induction. Top right panel shows turbidity measurements of Tau alone as negative control for LLPS detection. One‐way ANOVA with Tukey test for multiple comparison. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
Figure 3 Phosphorylation modulates Tau condensation and condensate affinity to surfaces Recombinant phosphorylated Tau expressed in insect cells (p‐Tau) de‐mixes in the presence of PEG but fails to coacervate with RNA (5 μM p‐Tau, 10% PEG). In the presence of PEG, polyA‐Cy3 co‐condensates with p‐Tau (arrowheads). Scale bar = 10 μm.Tau phospho‐variants generated by in vitro phosphorylation of non‐phosphorylated Tau from E. coli with different kinases (PKA, GSK3β, or Cdk5) show differences in phosphorylation at specific phospho‐site (pT205, pS262/pS356, pS324, and pS404) tested by Western blot. Non‐phosphorylated Tau and p‐Tau from insect cells are loaded for comparison. Position of tested p‐sites is indicated in the Tau sequence cartoon. Western blot for total Tau (Tau13 antibody) confirms equal loading (0.2 μg Tau per lane) of the Tau preparations.Phos‐tag gel shows degree and heterogeneity of phosphorylation in the different Tau phospho‐variants. Intensity along the lanes are plotted as colored lines next to gel image.Condensation of Tau phospho‐variants in the Tau:PEG, Tau:RNA (polyA), and pTau:PEG:RNA systems was determined by turbidity measurements for condensates in solution (absorption at 600 m) and by microscopy for surface attached condensates (surface coverage; top left). Turbidity measurements are shown on the left y‐axis (blue circles), and surface coverage in % on right y‐axis (black data points). Scale bars = 20 μm. N = 4 experiments, data points represent the average of 5 data points from turbidity measurements taken at 20 ± 5 min after LLPS induction. Top right panel shows turbidity measurements of Tau alone as negative control for LLPS detection. One‐way ANOVA with Tukey test for multiple comparison. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.CC BY-NC-ND · PubMed Central
Figure 4 Tubulin and RNA compete for condensed Tau Tubulin co‐condensation with Tau. Addition of tubulin‐Alexa647 (5 μM, 10:1 unlabeled:labeled tubulin) to a Tau‐Dy488 solution (25 μM with 1% Tau‐DyLight488) induces Tau condensation. Addition of tubulin to Tau:PEG, Tau:RNA (polyA‐Cy3), and Tau:PEG:RNA condensates leads to partitioning of tubulin into existing condensates and the occurrence of additional small Tau:tubulin condensates. Condensates with high tubulin‐HiLyte647 intensity have low polyA‐Cy3 intensity, and vice versa. In Tau:PEG:polyA condensates, individual condensates host RNA‐ and tubulin‐rich domains (arrowheads) at homogeneous Tau‐Dy488 intensity. Scale bars = 20 μm.Average intensity of RNA (top graph) normalized to Tau intensity in individual condensates. Addition of tubulin reduces the average amount of RNA in condensates. Intensity of tubulin (middle graph) normalized to Tau intensity in individual condensates. The presence of RNA increases the amount of tubulin in the condensates. Data shown as mean ± SD. Data +/− tubulin and +/− polyA were compared by two‐tailed Student’s t‐tests. Tubulin and RNA intensity correlate negatively on an individual condensate level (bottom graph). Line represent linear regression to all data. N = 3 experiments. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.Condensate size (are on surface) relative to tubulin (yellow data points) and RNA (pink data points) intensity. Smaller (new) condensates are high in tubulin, larger (old) condensates have more RNA. Lines represent linear regressions to tubulin (yellow) and RNA (pink) data.MT polymerization out of Tau:PEG and Tau:PEG:RNA condensates. Addition of tubulin‐Alexa647 (5 μM, 10:1 unlabeled:labeled tubulin) and GTP to Tau‐Dy488 (25 μM) diluted in BRB80 polymerization buffer induces MT polymerization. Images taken 30 min after addition of tubulin show thin partially curved MT bundles in Tau:PEG system, and thick bundles attached to persisting Tau:RNA condensates in the Tau:PEG:RNA system. Scale bar = 5 μm.Cross‐sectional intensity plots (line plots) of MT bundles differ between Tau:PEG (yellow/red line plots) and Tau:PEG:RNA (blue/green line plots) systems. Four line plots at random positions are shown per condition.Time course of MT polymerization out of Tau:PEG:RNA condensates. Scale bar = 20 μm. Box refers to panel G.Close‐up of Tau condensed phase connecting RNA condensate with MT bundles. Arrow indicates position of individual MT bundle positioned on top of RNA and underneath Tau in condensed phase. Scale bar = 5 μm.Nocodazole treatment (120 μM for 30 min) of MTs formed from Tau:PEG:RNA and Tau:PEG condensates. MT formed in absence of Tau shown as control. Graphs show surface coverage (%) of fibrillar MT structures before and after nocodazole treatment. N = 3 experiments, 3 images analyzed per experiment and condition. Data shown as mean ± SD. Two‐tailed Student’s t‐test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Scale bar = 20 μm.Attachment of Tau condensates to negatively charged (glutamate treated) glass surfaces. Tau:RNA and Tau:PEG:RNA condensates show pronounced surface wetting, whereas Tau:PEG condensates stay in solution or show partial wetting. p‐Tau:PEG condensates also wet the surfaces. Scale bars = 10 μm.
Figure 4 Tubulin and RNA compete for condensed Tau Tubulin co‐condensation with Tau. Addition of tubulin‐Alexa647 (5 μM, 10:1 unlabeled:labeled tubulin) to a Tau‐Dy488 solution (25 μM with 1% Tau‐DyLight488) induces Tau condensation. Addition of tubulin to Tau:PEG, Tau:RNA (polyA‐Cy3), and Tau:PEG:RNA condensates leads to partitioning of tubulin into existing condensates and the occurrence of additional small Tau:tubulin condensates. Condensates with high tubulin‐HiLyte647 intensity have low polyA‐Cy3 intensity, and vice versa. In Tau:PEG:polyA condensates, individual condensates host RNA‐ and tubulin‐rich domains (arrowheads) at homogeneous Tau‐Dy488 intensity. Scale bars = 20 μm.Average intensity of RNA (top graph) normalized to Tau intensity in individual condensates. Addition of tubulin reduces the average amount of RNA in condensates. Intensity of tubulin (middle graph) normalized to Tau intensity in individual condensates. The presence of RNA increases the amount of tubulin in the condensates. Data shown as mean ± SD. Data +/− tubulin and +/− polyA were compared by two‐tailed Student’s t‐tests. Tubulin and RNA intensity correlate negatively on an individual condensate level (bottom graph). Line represent linear regression to all data. N = 3 experiments. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.Condensate size (are on surface) relative to tubulin (yellow data points) and RNA (pink data points) intensity. Smaller (new) condensates are high in tubulin, larger (old) condensates have more RNA. Lines represent linear regressions to tubulin (yellow) and RNA (pink) data.MT polymerization out of Tau:PEG and Tau:PEG:RNA condensates. Addition of tubulin‐Alexa647 (5 μM, 10:1 unlabeled:labeled tubulin) and GTP to Tau‐Dy488 (25 μM) diluted in BRB80 polymerization buffer induces MT polymerization. Images taken 30 min after addition of tubulin show thin partially curved MT bundles in Tau:PEG system, and thick bundles attached to persisting Tau:RNA condensates in the Tau:PEG:RNA system. Scale bar = 5 μm.Cross‐sectional intensity plots (line plots) of MT bundles differ between Tau:PEG (yellow/red line plots) and Tau:PEG:RNA (blue/green line plots) systems. Four line plots at random positions are shown per condition.Time course of MT polymerization out of Tau:PEG:RNA condensates. Scale bar = 20 μm. Box refers to panel G.Close‐up of Tau condensed phase connecting RNA condensate with MT bundles. Arrow indicates position of individual MT bundle positioned on top of RNA and underneath Tau in condensed phase. Scale bar = 5 μm.Nocodazole treatment (120 μM for 30 min) of MTs formed from Tau:PEG:RNA and Tau:PEG condensates. MT formed in absence of Tau shown as control. Graphs show surface coverage (%) of fibrillar MT structures before and after nocodazole treatment. N = 3 experiments, 3 images analyzed per experiment and condition. Data shown as mean ± SD. Two‐tailed Student’s t‐test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Scale bar = 20 μm.Attachment of Tau condensates to negatively charged (glutamate treated) glass surfaces. Tau:RNA and Tau:PEG:RNA condensates show pronounced surface wetting, whereas Tau:PEG condensates stay in solution or show partial wetting. p‐Tau:PEG condensates also wet the surfaces. Scale bars = 10 μm.CC BY-NC-ND · PubMed Central
Figure EV3 MT polymerization from p‐Tau condensates Tau:polyA condensates do not form in MT polymerization buffer (BRB80). Scale bars = 20 and 5 μm.MT polymerization out of pTau:PEG and pTau:PEG:RNA condensates. Addition of tubulin (5 μM, with 10% tubulin HiLyte‐647) and GTP (1 mM) to p‐Tau‐Dy488 (25 μM with 1% p‐Tau‐DyLight488) diluted in BRB80 polymerization buffer induces MT polymerization out of condensates. pTau alone did not induce MT polymerization. pTau:PEG systems show thin MT bundles, and pTau:PEG condensates attach and wet onto MTs. In pTau:PEG:RNA systems, we observe thick bundles attached to persisting pTau:RNA condensates, similar to Tau:PEG:RNA systems. p‐Tau condensed phases extend along MT bundles beyond RNA condensed phase (indicated by dashed white line in inset), indicating preferred MT surface wetting by the Tau condensate. p‐Tau alone does not promote tubulin polymerization. p‐Tau was purified from insect cells. Images were taken 30 min after addition of tubulin. Scale bars = 20 and 5 μm.Attachment of Tau condensates to positively charged glass surfaces (amine treated TC‐dishes, Miltenyi). Tau:RNA and Tau:PEG:RNA condensates show surface wetting, whereas Tau:PEG condensates stay in solution or show partial wetting. Scale bars = 10 μm.
Figure EV3 MT polymerization from p‐Tau condensates Tau:polyA condensates do not form in MT polymerization buffer (BRB80). Scale bars = 20 and 5 μm.MT polymerization out of pTau:PEG and pTau:PEG:RNA condensates. Addition of tubulin (5 μM, with 10% tubulin HiLyte‐647) and GTP (1 mM) to p‐Tau‐Dy488 (25 μM with 1% p‐Tau‐DyLight488) diluted in BRB80 polymerization buffer induces MT polymerization out of condensates. pTau alone did not induce MT polymerization. pTau:PEG systems show thin MT bundles, and pTau:PEG condensates attach and wet onto MTs. In pTau:PEG:RNA systems, we observe thick bundles attached to persisting pTau:RNA condensates, similar to Tau:PEG:RNA systems. p‐Tau condensed phases extend along MT bundles beyond RNA condensed phase (indicated by dashed white line in inset), indicating preferred MT surface wetting by the Tau condensate. p‐Tau alone does not promote tubulin polymerization. p‐Tau was purified from insect cells. Images were taken 30 min after addition of tubulin. Scale bars = 20 and 5 μm.Attachment of Tau condensates to positively charged glass surfaces (amine treated TC‐dishes, Miltenyi). Tau:RNA and Tau:PEG:RNA condensates show surface wetting, whereas Tau:PEG condensates stay in solution or show partial wetting. Scale bars = 10 μm.CC BY-NC-ND · PubMed Central
Figure 5 Tau coacervates only percolate in the presence of PEG FRAP experiments of Tau:PEG, Tau:polyA, and Tau:PEG:polyA condensates at 1 h, 4 h, and 24 h after LLPS induction. Tau:PEG and Tau:PEG:polyA condensates show a decrease in recovery rate and mobile fraction over time. Tau:polyA coacervates maintain fast recovery and a large mobile fraction even after 24 h. Time until half maximum recovery: t(y1/2)polyA: 1 h ≈ 2 s, 4 h ≈ 3.5 s, 24 h ≈ 4.5 s; t(y1/2)PEG: 1 h ≈ 6 s, 4 h ≈ 8 s, 24 h ≈ 11 s; t(y1/2)PEG+polyA: 1 h ≈ 4.5 s, 4 h ≈ 9.5 s, 24 h ≈ 9 s). Scale bars = 5 μm for Tau:polyA and Tau:PEG, and 10 μm for Tau:PEG:polyA. Data presented as error bars (SEM), N = 15–20 condensates per condition and time point.Microscopy of Tau:PEG:polyA condensates (Tau‐Dylight488 and polyA‐Cy3) shows condensates with different Cy3 intensities. The bimodal fluorescence intensity distribution suggests two groups of condensates with high or low Cy3 fluorescence. Cy3‐intensities, normalized to Tau‐DyLight488 intensities in the same condensates, were independent of condensate size. Scale bar = 10 μm.FRAP of Tau‐Dylight488 and polyA‐Cy3 of 4 h‐old Tau:PEG:polyA condensates. The mean FRAP of polyA‐Cy3 is lower than of Tau‐Dy488 (left). Data shown as mean ± SEM, n = 24 condensates. FRAP data of individual condensates (bottom; #1 to #3) shows a decrease of Tau = Dy488 (middle) and polyA‐Cy3 (right) FRAP with an increase of polyA‐Cy3 (relative to Tau‐Dy488) in the condensates. Scale bar = 5 μm.
Figure 5 Tau coacervates only percolate in the presence of PEG FRAP experiments of Tau:PEG, Tau:polyA, and Tau:PEG:polyA condensates at 1 h, 4 h, and 24 h after LLPS induction. Tau:PEG and Tau:PEG:polyA condensates show a decrease in recovery rate and mobile fraction over time. Tau:polyA coacervates maintain fast recovery and a large mobile fraction even after 24 h. Time until half maximum recovery: t(y1/2)polyA: 1 h ≈ 2 s, 4 h ≈ 3.5 s, 24 h ≈ 4.5 s; t(y1/2)PEG: 1 h ≈ 6 s, 4 h ≈ 8 s, 24 h ≈ 11 s; t(y1/2)PEG+polyA: 1 h ≈ 4.5 s, 4 h ≈ 9.5 s, 24 h ≈ 9 s). Scale bars = 5 μm for Tau:polyA and Tau:PEG, and 10 μm for Tau:PEG:polyA. Data presented as error bars (SEM), N = 15–20 condensates per condition and time point.Microscopy of Tau:PEG:polyA condensates (Tau‐Dylight488 and polyA‐Cy3) shows condensates with different Cy3 intensities. The bimodal fluorescence intensity distribution suggests two groups of condensates with high or low Cy3 fluorescence. Cy3‐intensities, normalized to Tau‐DyLight488 intensities in the same condensates, were independent of condensate size. Scale bar = 10 μm.FRAP of Tau‐Dylight488 and polyA‐Cy3 of 4 h‐old Tau:PEG:polyA condensates. The mean FRAP of polyA‐Cy3 is lower than of Tau‐Dy488 (left). Data shown as mean ± SEM, n = 24 condensates. FRAP data of individual condensates (bottom; #1 to #3) shows a decrease of Tau = Dy488 (middle) and polyA‐Cy3 (right) FRAP with an increase of polyA‐Cy3 (relative to Tau‐Dy488) in the condensates. Scale bar = 5 μm.CC BY-NC-ND · PubMed Central
Figure EV4 Percolation and oligomerization of Tau condensates FRAP of 4 h‐old Tau:polyA condensates comparing different FRAP ROI sizes at similar condensate size (ROI1 versus ROI2) and full condensate bleach (ROI3). Left panel: FRAP data, Right panel: images showing ROI1,2,3 before and after bleach (at 0 s), and after recovery (at 40 s) the data were taken from. Scale bar = 10 μm.FRAP of tRNA and heparin induced Tau coacervates show maintained molecular diffusion even after 4 h and 24 h. Data presented as error bars (SD), N = 15–20 condensates per condition and time point.Tau:PEG:tRNA condensates show little FRAP of Tau‐Dy488 and tRNA‐Cy3 after 4 h. Data shown as mean±SD. N = 25 condensates per condition.SDS–PAGE of Tau condensate sedimentation assay. Tau:RNA and Tau:heparin condensates do not enrich in the dense phase (P) upon centrifugation at 16,000 g for 15 min at room temperature, hence have a lower density than Tau:PEG and Tau:PEG:polyA condensates, for which the dense (P) and light (S) phase can be separated by centrifugation.Semi‐Native PAGE of Tau incubated in LLPS (PEG, tRNALLPS, heparinLLPS) or aggregation (tRNAAGG, heparinAGG) conditions for 1 h, 24 h, or 72 h. Higher molecular weight Tau assemblies are prominent in aggregation conditions after 24 h, and in all conditions after 72 h, even in the Tau only control. Tau dimers are present already after 1 h in all samples.SDS–PAGE and Semi‐Native PAGE of 5 μM Tau in condensation and aggregation conditions incubated for 24 h show the formation of SDS‐stable Tau dimers in condensate.Dot blot analysis of 24 h‐old Tau LLPS and aggregation samples indicates an increase of oligomeric Tau (2B10 antibody; P = 0.043 for tRNALLPS) in Tau LLPS samples containing tRNA or PEG compared to Tau only treatment. N = 3 independent experiments. Data shown as mean ± SD. One‐way ANOVA with Holmes‐Sidak test for multiple comparison. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
Figure EV4 Percolation and oligomerization of Tau condensates FRAP of 4 h‐old Tau:polyA condensates comparing different FRAP ROI sizes at similar condensate size (ROI1 versus ROI2) and full condensate bleach (ROI3). Left panel: FRAP data, Right panel: images showing ROI1,2,3 before and after bleach (at 0 s), and after recovery (at 40 s) the data were taken from. Scale bar = 10 μm.FRAP of tRNA and heparin induced Tau coacervates show maintained molecular diffusion even after 4 h and 24 h. Data presented as error bars (SD), N = 15–20 condensates per condition and time point.Tau:PEG:tRNA condensates show little FRAP of Tau‐Dy488 and tRNA‐Cy3 after 4 h. Data shown as mean±SD. N = 25 condensates per condition.SDS–PAGE of Tau condensate sedimentation assay. Tau:RNA and Tau:heparin condensates do not enrich in the dense phase (P) upon centrifugation at 16,000 g for 15 min at room temperature, hence have a lower density than Tau:PEG and Tau:PEG:polyA condensates, for which the dense (P) and light (S) phase can be separated by centrifugation.Semi‐Native PAGE of Tau incubated in LLPS (PEG, tRNALLPS, heparinLLPS) or aggregation (tRNAAGG, heparinAGG) conditions for 1 h, 24 h, or 72 h. Higher molecular weight Tau assemblies are prominent in aggregation conditions after 24 h, and in all conditions after 72 h, even in the Tau only control. Tau dimers are present already after 1 h in all samples.SDS–PAGE and Semi‐Native PAGE of 5 μM Tau in condensation and aggregation conditions incubated for 24 h show the formation of SDS‐stable Tau dimers in condensate.Dot blot analysis of 24 h‐old Tau LLPS and aggregation samples indicates an increase of oligomeric Tau (2B10 antibody; P = 0.043 for tRNALLPS) in Tau LLPS samples containing tRNA or PEG compared to Tau only treatment. N = 3 independent experiments. Data shown as mean ± SD. One‐way ANOVA with Holmes‐Sidak test for multiple comparison. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.CC BY-NC-ND · PubMed Central
Figure 6 Tau coacervates and condensates produce Tau species with seeding potential SDS–PAGE of Tau incubated in LLPS (PEG, tRNALLPS, PEG:tRNALLPS) or aggregation (hepAGG) conditions for 1 h, 24 h, or 72 h. SDS‐stable Tau dimers and oligomers species evolve in LLPS and hepAGG samples after 24 h. After 72 h, all conditions contain SDS‐stable Tau oligomers and dimers.Thioflavin‐T assay for the enhancement of in vitro Tau aggregation by aged Tau condensates. TauΔK280 (10 µM) incubated with heparin (molar ratio Tau:heparin = 4:1; [TauΔK280]AGG) was used as baseline condition for Tau aggregation. Addition of 24 h‐old or 72 h‐old Tau LLPS samples (9.2 µg/ml = 0.2 µM Tau) produced an increase in TauΔK280 aggregation for all LLPS conditions (PEG, tRNA, PEG:tRNA) and hepAGG. Addition of Tau only did not affect TauΔK280 aggregation. Data shown as mean ± SD, n = 3. One‐way ANOVA with Tukey test for multiple comparison for values at 30 h. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.HEK sensor cell Tau aggregation assay. Cells were treated with 24 h‐old condensates or aggregates (1.2 µMTau in the presence of 0.8% lipofectamine). The number of Tau aggregates was determined from low magnification (20×) images taken in the green channel and normalized to the number of cells (DAPI). Scale bar = 100 μm.Quantification of seeding assay in (C). Data were collected from three images of 3 wells for each condition. Numbers in the graph give the number of analyzed cells per condition. Data shown as mean ± SD, N = 3 experiments, 1–3 images per experiment and condition. One‐way ANOVA with Tukey test for multiple comparison. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.Cell toxicity (LDH assay) induced by treatment of HEK sensor cells with condensates or aggregates. N = 3 experiments. Data shown as mean ± SD. One‐way ANOVA with Tukey test for multiple comparison. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
Figure 6 Tau coacervates and condensates produce Tau species with seeding potential SDS–PAGE of Tau incubated in LLPS (PEG, tRNALLPS, PEG:tRNALLPS) or aggregation (hepAGG) conditions for 1 h, 24 h, or 72 h. SDS‐stable Tau dimers and oligomers species evolve in LLPS and hepAGG samples after 24 h. After 72 h, all conditions contain SDS‐stable Tau oligomers and dimers.Thioflavin‐T assay for the enhancement of in vitro Tau aggregation by aged Tau condensates. TauΔK280 (10 µM) incubated with heparin (molar ratio Tau:heparin = 4:1; [TauΔK280]AGG) was used as baseline condition for Tau aggregation. Addition of 24 h‐old or 72 h‐old Tau LLPS samples (9.2 µg/ml = 0.2 µM Tau) produced an increase in TauΔK280 aggregation for all LLPS conditions (PEG, tRNA, PEG:tRNA) and hepAGG. Addition of Tau only did not affect TauΔK280 aggregation. Data shown as mean ± SD, n = 3. One‐way ANOVA with Tukey test for multiple comparison for values at 30 h. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.HEK sensor cell Tau aggregation assay. Cells were treated with 24 h‐old condensates or aggregates (1.2 µMTau in the presence of 0.8% lipofectamine). The number of Tau aggregates was determined from low magnification (20×) images taken in the green channel and normalized to the number of cells (DAPI). Scale bar = 100 μm.Quantification of seeding assay in (C). Data were collected from three images of 3 wells for each condition. Numbers in the graph give the number of analyzed cells per condition. Data shown as mean ± SD, N = 3 experiments, 1–3 images per experiment and condition. One‐way ANOVA with Tukey test for multiple comparison. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.Cell toxicity (LDH assay) induced by treatment of HEK sensor cells with condensates or aggregates. N = 3 experiments. Data shown as mean ± SD. One‐way ANOVA with Tukey test for multiple comparison. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.CC BY-NC-ND · PubMed Central
Figure EV5 Oligomerization and seeding of pro‐aggregant TauΔK280 condensates A, BSDS–PAGE (A) and SemiNative PAGE (B) of TauΔK280 incubated in LLPS (PEG, tRNALLPS, heparinLLPS) or aggregation (tRNAAGG, heparinAGG) conditions for 1 h, 24 h, or 72 h.CNative PAGE and SDS–PAGE of 5 μM TauΔK280 in condensation and aggregation conditions incubated for 24 h. Even for low amounts of TauΔK280, PEG‐, RNA‐ and heparin‐induced condensates lead to the formation of SDS‐stable Tau dimers.DDot blot analysis of TauΔK280 oligomers (2B10 antibody) in 24 h‐old condensation and aggregation samples shows oligomeric Tau in aggregation conditions compared to Tau only treatment. N = 3 independent experiments. Data shown as mean ± SEM, one‐way ANOVA with Holmes‐Sidak post‐test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.EThioflavin‐T assay for the enhancement of in vitro Tau aggregation by TauΔK280 condensates. TauΔK280 (10 µM) incubated with heparin (molar ratio Tau:heparin = 4:1; [TauΔK280]AGG) was used as baseline condition for Tau aggregation. Addition of 24 h‐old or 72 h‐old TauΔK280 LLPS samples (9.2 μg/ml, molar ratio TauΔK280:condensates = 50:1) to the reaction showed a significant increase of TauΔK280 aggregation for all LLPS conditions (PEGLLPS, tRNALLPS, PEG:tRNALLPS), which was comparable to the effect of pre‐aggregated TauΔK280 (hepAGG). Addition of TauΔK280 only showed no effect on TauΔK280 aggregation. Data shown as mean ± SD, N = 3 experiments, with 3 technical replicates per experiment. One‐way ANOVA with Tukey post‐test for multiple comparison for values at 30 h. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
Figure EV5 Oligomerization and seeding of pro‐aggregant TauΔK280 condensates A, BSDS–PAGE (A) and SemiNative PAGE (B) of TauΔK280 incubated in LLPS (PEG, tRNALLPS, heparinLLPS) or aggregation (tRNAAGG, heparinAGG) conditions for 1 h, 24 h, or 72 h.CNative PAGE and SDS–PAGE of 5 μM TauΔK280 in condensation and aggregation conditions incubated for 24 h. Even for low amounts of TauΔK280, PEG‐, RNA‐ and heparin‐induced condensates lead to the formation of SDS‐stable Tau dimers.DDot blot analysis of TauΔK280 oligomers (2B10 antibody) in 24 h‐old condensation and aggregation samples shows oligomeric Tau in aggregation conditions compared to Tau only treatment. N = 3 independent experiments. Data shown as mean ± SEM, one‐way ANOVA with Holmes‐Sidak post‐test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.EThioflavin‐T assay for the enhancement of in vitro Tau aggregation by TauΔK280 condensates. TauΔK280 (10 µM) incubated with heparin (molar ratio Tau:heparin = 4:1; [TauΔK280]AGG) was used as baseline condition for Tau aggregation. Addition of 24 h‐old or 72 h‐old TauΔK280 LLPS samples (9.2 μg/ml, molar ratio TauΔK280:condensates = 50:1) to the reaction showed a significant increase of TauΔK280 aggregation for all LLPS conditions (PEGLLPS, tRNALLPS, PEG:tRNALLPS), which was comparable to the effect of pre‐aggregated TauΔK280 (hepAGG). Addition of TauΔK280 only showed no effect on TauΔK280 aggregation. Data shown as mean ± SD, N = 3 experiments, with 3 technical replicates per experiment. One‐way ANOVA with Tukey post‐test for multiple comparison for values at 30 h. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.CC BY-NC-ND · PubMed Central
Figure EV6 Tau accumulation “seeded” by 24 h‐old Tau condensates HEK sensor cells treated with 1 h‐ and 24 h‐old Tau condensates. No seeding occurs in 1 h‐old LLPS preparations, whereas 24 h‐old condensates efficiently induce aggregation. Notably, a minor number of aggregates (inset) could be found already in 1 h‐old Tau:PEG:tRNA condensates, indicating a rapid production of seeding competent Tau in these condensates.Confocal images of HEK sensor cells treated with Tau condensates (top) or TauΔK280 aggregates (bottom). Some nuclei are outlined with white dashed lines for better visualization. Scale bars = 20 μm in overview, and 10 μm in close‐ups. Asterisks indicate position of cell nucleus, dashed lines indicate the nucleus border.Tau sensor neurons treated with AD brain lysate show Tau accumulations (CYT, NUC, and NE) similar to HEK sensor cells. Mouse primary hippocampal neurons were transduced with AAV TauRDP301S‐CFP.P2a.TauRDP301S‐YFP and three days later treated with AD brain TBS extract. Tau accumulations (arrows) were imaged 6 days later on DIV12. Non‐treated neurons did not develop Tau accumulations. Scale bars = 10 μm.
Figure EV6 Tau accumulation “seeded” by 24 h‐old Tau condensates HEK sensor cells treated with 1 h‐ and 24 h‐old Tau condensates. No seeding occurs in 1 h‐old LLPS preparations, whereas 24 h‐old condensates efficiently induce aggregation. Notably, a minor number of aggregates (inset) could be found already in 1 h‐old Tau:PEG:tRNA condensates, indicating a rapid production of seeding competent Tau in these condensates.Confocal images of HEK sensor cells treated with Tau condensates (top) or TauΔK280 aggregates (bottom). Some nuclei are outlined with white dashed lines for better visualization. Scale bars = 20 μm in overview, and 10 μm in close‐ups. Asterisks indicate position of cell nucleus, dashed lines indicate the nucleus border.Tau sensor neurons treated with AD brain lysate show Tau accumulations (CYT, NUC, and NE) similar to HEK sensor cells. Mouse primary hippocampal neurons were transduced with AAV TauRDP301S‐CFP.P2a.TauRDP301S‐YFP and three days later treated with AD brain TBS extract. Tau accumulations (arrows) were imaged 6 days later on DIV12. Non‐treated neurons did not develop Tau accumulations. Scale bars = 10 μm.CC BY-NC-ND · PubMed Central
Figure 7 Nuclear envelope Tau accumulations seeded by condensates and in AD brain Confocal microscopy of HEK sensor cells treated with 24 h‐old Tau condensates (Tau:tRNA) shows different types of intracellular Tau‐CFP/YFP accumulations: spherical intranuclear Tau (NUC), bright cytoplasmic accumulations (CYT) adjacent to the nucleus, and small, dimmer accumulations right at and along the nuclear envelope (NE). Cells treated with Tau aggregates (hepAGG) had mainly large cytoplasmic (CYT) Tau inclusions. The different Tau accumulations coexisted in most cells, and also more than one accumulation of the same type could occur. Nuclei were stained with propidium iodide (PI, pink). Scale bars = 10 μm.Fluorescence intensity of NE and CYT Tau compared to soluble Tau in the cytosol of cells without accumulations. N = 3 experiments, and 5–6 images per experiment. Data shown as mean ± SD. One‐way ANOVA with Tukey test for multiple comparison. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.Immunolabeling of lamin‐B1 and nuclear pore complexes (NPC) shows the location of CYT and NE Tau in relation to the nuclear envelope in seeded HEK sensor cells. White arrows and boxes show position and width of cross‐sectional profiles shown in (D). Scale bar = 10 μm.Cross‐sectional profiles show the localization of Tau (green) in CYT accumulations adjacent to the nucleus in the cytosol, and the colocalization of NE Tau with lamin (pink) and NPCs (blue). Notably, CYT‐1 also contains NPC signal. Dashed vertical lines indicate x‐positions of highest intensity values for respective intensity profiles.Seeded HEK sensor cells stained with the amyloid dye (Amytracker‐680). Scale bar = 10 μm.HEK sensor cells treated with AD brain TBS extract show similar CYT, NUC, and NE accumulations as Tau condensate treated cells. Arrowheads and arrows indicate positions of respective Tau accumulation types. Scale bars = 10 μm.Immunolabeling of phosphorylated Tau (PHF1 = pS396/404, AT8 = pS199/pS202/pT205, AT180 = pT231) in AD brain section. Both 3,3′‐Diaminobenzidine (DAB) and fluorescence labeling show Tau accumulating in a layer at the nuclear envelope (NE) in non‐tangle neurons with or without cytosolic (CYT) Tau accumulations. Control brain does not show Tau labeling in neuronal cell bodies. Scale bars = 20 and 10 μm, as indicated.High resolution confocal images show colocalization (arrowheads) of small NE Tau clusters with nuclear pore proteins (NPC marker, POM121). Scale bars = 10 μm.Reduced nucleo‐cytoplasmic Ran gradient (nuclear/cytosolic Ran intensity) in HEK sensor cells having CYT Tau (CYT), or NE Tau (NE), compared to non‐treated (NT) or control treated (lipo) sensor cells. Wild‐type (WT) HEK cell Ran gradient was analyzed for the effect of TauRDP301S expression in general. N = 3 experiments, 7–10 cells analyzed per experiment and condition. Data shown as mean ± SD. One‐way ANOVA with Tukey test for multiple comparison. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.Frequency of Tau accumulation types observed in sensor cells treated with Tau aggregates (hepAGG), AD lysate, or Tau condensates. In cells having Tau accumulations we determined which accumulation types were present, regardless of the co‐presence of other accumulation types and if multiple accumulations of the same type were present. n = number of cells with accumulations analyzed, resulted from the actual number of cells analyzed in the three replicates.
Figure 7 Nuclear envelope Tau accumulations seeded by condensates and in AD brain Confocal microscopy of HEK sensor cells treated with 24 h‐old Tau condensates (Tau:tRNA) shows different types of intracellular Tau‐CFP/YFP accumulations: spherical intranuclear Tau (NUC), bright cytoplasmic accumulations (CYT) adjacent to the nucleus, and small, dimmer accumulations right at and along the nuclear envelope (NE). Cells treated with Tau aggregates (hepAGG) had mainly large cytoplasmic (CYT) Tau inclusions. The different Tau accumulations coexisted in most cells, and also more than one accumulation of the same type could occur. Nuclei were stained with propidium iodide (PI, pink). Scale bars = 10 μm.Fluorescence intensity of NE and CYT Tau compared to soluble Tau in the cytosol of cells without accumulations. N = 3 experiments, and 5–6 images per experiment. Data shown as mean ± SD. One‐way ANOVA with Tukey test for multiple comparison. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.Immunolabeling of lamin‐B1 and nuclear pore complexes (NPC) shows the location of CYT and NE Tau in relation to the nuclear envelope in seeded HEK sensor cells. White arrows and boxes show position and width of cross‐sectional profiles shown in (D). Scale bar = 10 μm.Cross‐sectional profiles show the localization of Tau (green) in CYT accumulations adjacent to the nucleus in the cytosol, and the colocalization of NE Tau with lamin (pink) and NPCs (blue). Notably, CYT‐1 also contains NPC signal. Dashed vertical lines indicate x‐positions of highest intensity values for respective intensity profiles.Seeded HEK sensor cells stained with the amyloid dye (Amytracker‐680). Scale bar = 10 μm.HEK sensor cells treated with AD brain TBS extract show similar CYT, NUC, and NE accumulations as Tau condensate treated cells. Arrowheads and arrows indicate positions of respective Tau accumulation types. Scale bars = 10 μm.Immunolabeling of phosphorylated Tau (PHF1 = pS396/404, AT8 = pS199/pS202/pT205, AT180 = pT231) in AD brain section. Both 3,3′‐Diaminobenzidine (DAB) and fluorescence labeling show Tau accumulating in a layer at the nuclear envelope (NE) in non‐tangle neurons with or without cytosolic (CYT) Tau accumulations. Control brain does not show Tau labeling in neuronal cell bodies. Scale bars = 20 and 10 μm, as indicated.High resolution confocal images show colocalization (arrowheads) of small NE Tau clusters with nuclear pore proteins (NPC marker, POM121). Scale bars = 10 μm.Reduced nucleo‐cytoplasmic Ran gradient (nuclear/cytosolic Ran intensity) in HEK sensor cells having CYT Tau (CYT), or NE Tau (NE), compared to non‐treated (NT) or control treated (lipo) sensor cells. Wild‐type (WT) HEK cell Ran gradient was analyzed for the effect of TauRDP301S expression in general. N = 3 experiments, 7–10 cells analyzed per experiment and condition. Data shown as mean ± SD. One‐way ANOVA with Tukey test for multiple comparison. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.Frequency of Tau accumulation types observed in sensor cells treated with Tau aggregates (hepAGG), AD lysate, or Tau condensates. In cells having Tau accumulations we determined which accumulation types were present, regardless of the co‐presence of other accumulation types and if multiple accumulations of the same type were present. n = number of cells with accumulations analyzed, resulted from the actual number of cells analyzed in the three replicates.CC BY-NC-ND · PubMed Central
Figure 8 FRAP and FLIM reveal differences in molecular packing in Tau accumulations FRAP of Tau accumulations show no recovery of NUC Tau (NUC: <10% recovery; t1/2 ≈ 11 s), ~20% recovery of CYT and NE Tau with a faster recovery rate for NE Tau (NE: t1/2 ≈ 2.5 s; CYT: t1/2 ≈7 s). N = 15–20 Tau accumulations per type. Data shown as mean ± SD.Example FRAP image series for cytosol and the different Tau accumulation types in HEK sensor cells treated with AD brain lysate (pre‐bleach = −2 s, immediately after bleaching = 0 s, after recovery = 40 s). Arrows indicate position of bleached ROIs. Scale bars = 5 or 10 µm as indicated.High‐resolution CFP FLIM of HEK sensor cells expressing TauRDP301L‐CFP/YFP and treated with Tau condensates. FRET between CFP and YFP induces CFP‐lifetime reduction in densely packed TauRDP301L‐CFP/YFP accumulations. Fit‐free analysis of the lifetime distribution in phasor plots enables the assignment of distinct lifetime clusters (circles in phasor plot) to different Tau accumulation types. LT‐2 = NE Tau, LT‐3 = CYT, and NUC Tau, LT‐4 = fully aggregated CYT Tau. LT‐1 corresponds to soluble TauRD‐CFP in the cytosol of cells without accumulations. Asterisks indicate position of individual cell nuclei. Scale bars = 10 μm.CFP‐FLIM of seeded HEK sensor cells expressing only TauRDP301L‐CFP. The dense packing of TauRD‐CFP in Tau accumulations induces fluorophore quenching that can be detected as decrease in CFP‐lifetime. Fit‐free analysis of lifetime phasor plots enables the assignment of distinct lifetime clusters (circles in phasor plot) to different Tau accumulation types. LT‐2 = NE Tau, LT‐3 = CYT and NUC Tau, and LT‐1 = soluble TauRD‐CFP in the cytosol of cells without accumulations. Graphs show lifetimes of CFP‐lifetime components assigned to soluble, NE, and CYT + NUC Tau in 6 different images (left graph, RM one‐way ANOVA with Tukey post‐test) and lifetime reduction relative to LT‐1 in each respective image (right graph, one‐way ANOVA with Tukey post‐test.). N = 2 experiments, 2–3 cells measured per experiments. Data shown as mean ± SD. Scale bars = 10 μm. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.FLIM of 1 h‐ and 24 h‐old CFP‐Tau (full‐length) condensates and heparin‐induced CFP‐Tau aggregates. Lifetime components are indicated in phasor plots. Graph shows lifetime reduction in condensates and aggregates relative to respective lifetime of soluble CFP‐Tau. N = 2 experiments, 2–3 cells measured per experiments. Data shown as mean ± SD. One‐way ANOVA with Tukey post‐test. Scale bar = 5 μm. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Bottom right: Schematic of lifetime quenching in condensates and aggregates.
Figure 8 FRAP and FLIM reveal differences in molecular packing in Tau accumulations FRAP of Tau accumulations show no recovery of NUC Tau (NUC: <10% recovery; t1/2 ≈ 11 s), ~20% recovery of CYT and NE Tau with a faster recovery rate for NE Tau (NE: t1/2 ≈ 2.5 s; CYT: t1/2 ≈7 s). N = 15–20 Tau accumulations per type. Data shown as mean ± SD.Example FRAP image series for cytosol and the different Tau accumulation types in HEK sensor cells treated with AD brain lysate (pre‐bleach = −2 s, immediately after bleaching = 0 s, after recovery = 40 s). Arrows indicate position of bleached ROIs. Scale bars = 5 or 10 µm as indicated.High‐resolution CFP FLIM of HEK sensor cells expressing TauRDP301L‐CFP/YFP and treated with Tau condensates. FRET between CFP and YFP induces CFP‐lifetime reduction in densely packed TauRDP301L‐CFP/YFP accumulations. Fit‐free analysis of the lifetime distribution in phasor plots enables the assignment of distinct lifetime clusters (circles in phasor plot) to different Tau accumulation types. LT‐2 = NE Tau, LT‐3 = CYT, and NUC Tau, LT‐4 = fully aggregated CYT Tau. LT‐1 corresponds to soluble TauRD‐CFP in the cytosol of cells without accumulations. Asterisks indicate position of individual cell nuclei. Scale bars = 10 μm.CFP‐FLIM of seeded HEK sensor cells expressing only TauRDP301L‐CFP. The dense packing of TauRD‐CFP in Tau accumulations induces fluorophore quenching that can be detected as decrease in CFP‐lifetime. Fit‐free analysis of lifetime phasor plots enables the assignment of distinct lifetime clusters (circles in phasor plot) to different Tau accumulation types. LT‐2 = NE Tau, LT‐3 = CYT and NUC Tau, and LT‐1 = soluble TauRD‐CFP in the cytosol of cells without accumulations. Graphs show lifetimes of CFP‐lifetime components assigned to soluble, NE, and CYT + NUC Tau in 6 different images (left graph, RM one‐way ANOVA with Tukey post‐test) and lifetime reduction relative to LT‐1 in each respective image (right graph, one‐way ANOVA with Tukey post‐test.). N = 2 experiments, 2–3 cells measured per experiments. Data shown as mean ± SD. Scale bars = 10 μm. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.FLIM of 1 h‐ and 24 h‐old CFP‐Tau (full‐length) condensates and heparin‐induced CFP‐Tau aggregates. Lifetime components are indicated in phasor plots. Graph shows lifetime reduction in condensates and aggregates relative to respective lifetime of soluble CFP‐Tau. N = 2 experiments, 2–3 cells measured per experiments. Data shown as mean ± SD. One‐way ANOVA with Tukey post‐test. Scale bar = 5 μm. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Bottom right: Schematic of lifetime quenching in condensates and aggregates.CC BY-NC-ND · PubMed Central
Figure 9 Tau condensation in different environments and condensate‐induced cellular Tau accumulations Tau coacervation with RNA (1) ‐ or other polyanions – without PEG produces non‐percolating condensates that are instable at physiological ion concentrations. Phosphorylation (pTau) inhibits coacervation of Tau. Crowding‐induced condensates (2) are stable at physiological salt concentrations and percolate into gel‐like condensates within 24 h. Tau:PEG condensates promote MT growth but disassemble during the polymerization process. Condensation of Tau with RNA in the presence of molecular crowding (3), a condition that likely exists in the cytosol, produces condensates that are stable at physiological ion concentrations and also percolate into gel‐like condensates. Tau:PEG:RNA condensates strongly promote MT bundling, and they persist as condensed phases attached to MT bundles, thereby inducing a distinct spatial organizing of MT bundles. Tau:PEG:RNA condensates can function as “liquid matrix” that can host different otherwise not compatible Tau binding partners, such as RNA and tubulin. Regardless of their physiochemistry and origin, all condensates induce the formation of SDS‐stable Tau dimers and oligomers with time, and efficiently induce Tau aggregation in sensor cell and neuron models. Tau condensates induce spherical accumulations in the nucleus (NUC), large aggregates in the cytosol (CYT) adjacent to the nucleus, and small granular Tau condensates at the nuclear envelope (NE), which in part colocalize with nucleoporins and impair the nucleocytoplasmic transport. In contrast, heparin‐induced Tau aggregates mostly seed CYT accumulations.
Figure 9 Tau condensation in different environments and condensate‐induced cellular Tau accumulations Tau coacervation with RNA (1) ‐ or other polyanions – without PEG produces non‐percolating condensates that are instable at physiological ion concentrations. Phosphorylation (pTau) inhibits coacervation of Tau. Crowding‐induced condensates (2) are stable at physiological salt concentrations and percolate into gel‐like condensates within 24 h. Tau:PEG condensates promote MT growth but disassemble during the polymerization process. Condensation of Tau with RNA in the presence of molecular crowding (3), a condition that likely exists in the cytosol, produces condensates that are stable at physiological ion concentrations and also percolate into gel‐like condensates. Tau:PEG:RNA condensates strongly promote MT bundling, and they persist as condensed phases attached to MT bundles, thereby inducing a distinct spatial organizing of MT bundles. Tau:PEG:RNA condensates can function as “liquid matrix” that can host different otherwise not compatible Tau binding partners, such as RNA and tubulin. Regardless of their physiochemistry and origin, all condensates induce the formation of SDS‐stable Tau dimers and oligomers with time, and efficiently induce Tau aggregation in sensor cell and neuron models. Tau condensates induce spherical accumulations in the nucleus (NUC), large aggregates in the cytosol (CYT) adjacent to the nucleus, and small granular Tau condensates at the nuclear envelope (NE), which in part colocalize with nucleoporins and impair the nucleocytoplasmic transport. In contrast, heparin‐induced Tau aggregates mostly seed CYT accumulations.CC BY-NC-ND · PubMed Central

Verification

Claims the paper makes

0 Valid · 0 Not valid · 54 Not checked · 252 Own results and interpretations

146 more claims

Create a free account to see all 306 claims and to verify the 68 references still open.

See all claims — free account
  • It may state two separate claims as one. 117
  • Probably not a substantial claim. 51
  • The quote only partly states this claim. 2

Create a free account to see these claims.

Create a free account to see these assumptions.

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

5 of 6 · the rest aren’t loaded

1 more finding

Create a free account to see every finding of this check.

See all findings — free account

Citations

Cites

Cited by

See all 106

Related

See all

Metadata

Authors
Hochmair, J., Exner, C., Franck, M., Dominguez‐Baquero, A., Diez, L., Brognaro, H., Kraushar, M. L., Mielke, T., Radbruch, H., Kaniyappan, S., Falke, S., Mandelkow, E., Betzel, C., & Wegmann, S.
Affiliations
  • German Center for Neurodegenerative Diseases (DE)
  • Universität Hamburg (DE)
  • Max Planck Institute for Molecular Genetics (DE)
  • Charité – Universitätsmedizin Berlin (DE)
  • University of Bonn (DE)
Corresponding author
Wegmann, Susanne
ORCID
Venue
The EMBO Journal
Publisher
Springer Science and Business Media LLC
Published
2022-03-17
Volume
41
Issue
11
Pages
e108882-e108882
ISSN
  • 0261-4189 (ISSN-L)
  • 1460-2075
DOI
10.15252/embj.2021108882
PMC
PMC9156969
OpenAlex
W4221061896
Cited by
106
References
94
Citations by year
2026: 38 · 2025: 25 · 2024: 12 · 2023: 22 · 2022: 9
FWCI
5.78
Type
journal article
Language
en
Open access
Licence
Locations
  • The EMBO Journal (publishedVersion, cc-by-nc-nd)
  • bioRxiv (Cold Spring Harbor Laboratory) (submittedVersion)
  • PubMed (publishedVersion)
  • PubMed Central (publishedVersion, cc-by-nc-nd)
  • DZNE Pub (submittedVersion, other-oa)
Funders
  • Deutsche Forschungsgemeinschaft (402723784, EXC 2056)
  • Deutsches Zentrum für Neurodegenerative Erkrankungen
  • Gemeinnützige Hertie‐Stiftung
  • BrightFocus Foundation
Topics
  • Life Sciences › Neuroscience › Cellular and Molecular Neuroscience › Alzheimer's disease research and treatments
  • Health Sciences › Medicine › Clinical Neurology › Amyotrophic Lateral Sclerosis Research
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Cell Biology › Microtubule and mitosis dynamics
Keywords
tau protein, biomolecular condensates, liquid-liquid phase separation, macromolecular crowding, microtubule bundling, microtubule assembly, tau aggregation, nuclear transport, fluorescence lifetime imaging, molecular packing density, neurodegenerative diseases
MeSH terms
Humans, Microtubules, Neurons, Protein Binding, RNA, tau Proteins, Neurodegenerative Diseases

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-11-13, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-11-13 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.