METHOD OF PREPARATION OF AMYLOIDOGENIC PROTEIN AGGREGATES AND

USES THEREOF

Field of the Invention

The present invention pertains generally to the field of in vitro aggregation amyloidogenic proteins that need a co-factor, in particular Tau proteins, and the use of the resulting aggregates, fibrils, soluble and insoluble oligomers and uses thereof.

Background of the Invention

Tau protein aggregation is implicated in the pathogenesis of several neurodegenerative diseases collectively known as Tauopathies.

The microtubule-binding protein (MAP) Tau is an intrinsically disordered protein (IDP) that is most prominently associated with the dynamic regulation and stabilization of cytoskeletal and mitotic microtubules (Drubin, 1986, J Cell Biol 103, 2739, doi: 10.1083/jcb.103.6.2739). In neurons, Tau is also important for regulating axon outgrowth and maintaining axonal transport and cytoskeletal integrity (Dixit et al., 2008, Science 319, 1086-1089). However, factors such as post- translational modifications (PTMs) (Ding et al., 2006, Journal of Biological Chemistry 281, 19107- 19114), mutations in the protein sequence (Barghorn et al., 2000, Biochemistry 39, 11714-11721), interaction with other proteins (Baudier & Cole, 1988, Journal of Biological Chemistry 263, 5876- 5883; Carlier et al, 1984, Biochimie, 66, 305-311) and changes to the biochemistry of its surrounding environment, such as pH or the presence of drugs (Samsonov, 2004, J Cell Sci., 117, 6129-6141, doi: 10.1242/jcs.01531), may result in the lowered affinity, weaker interaction or full dissociation of Tau from microtubules (Venkatramani et al, 2019, International journal of biological macromolecules, 133, 473-483, doi:https://doi.org/l 0.1016/i.iibiomac.2019.04.120). Tau may then accumulate and aggregate into higher molecular weight species, such as fibrils associated with pathology. Increasing evidence points to Tau aggregation and PTMs as central events in the pathogenesis of Alzheimer’s disease (AD) and Tauopathies, events that investigators strive to faithfully model in the laboratory (Limorenko, 2021, Neurobiol Dis 161, 105536).

In addition to amyloid plaques composed of P-amyloid, a classic hallmark of AD, another prominent pathological feature of AD is hyperphosphorylated Tau which is found in neuronal cell bodies or neurites in the form of paired helical filaments (PHFs) and straight filaments (SFs) (Golde et al, 2022, Mol Neurodegener, 17, 18, doi: 10.1186/sl 3024-022-00523-1).

Tau aggregates and fibrillar structures are also found in the brain of individuals afflicted by other neurodegenerative diseases (NDs), collectively known as Tauopathies, which include Pick’s disease (PiD) and progressive supranuclear palsy (PSP) (Murray etal., 2017, Alzheimers Res Ther., 6, 1).

Tau exists as six isoforms in the human central nervous system, designated 4R2N, 4R1N, 4R0N, 3R2N, 3R1N, and 3R0N. Where the Tau isoform compositions of the Tau fibrils are known (Lee, 2001, Annu Rev Neurosci., 24, 1121-1159), Tauopathies are classified into predominantly 3R (i.e., PiD), predominantly 4R (i.e,. PSP), or mixed (3R + 4R; i.e., AD).

Full-length Tau isoforms are highly soluble and notoriously resistant to aggregation on their own, in contrast to other amyloid-forming proteins, such as a-synuclein (Horvath, 2018, in Amyloid Proteins, 73-83 (Springer), P-amyloid peptide (Sasanian et al., 2020, Biomolecules 10, 924) or amylin (Azzam, 2018, Molecular Pharmaceutics, 15, 2098-2106, doi:10.1021/acs.molpharmaceut.7b01009) which misfold, aggregate and form amyloid fibrils simply by incubation at 37°C. The molecular and cellular factors that trigger Tau misfolding and aggregation, and drive the Tau fibrillization processes remain unclear. Therefore, to study Tau fibrillization in vitro, investigators have adopted Tau aggregation systems using various negatively- charged co-factor molecules or protein modifications, such as truncations or phosphorylation, which are used to induce or accelerate the full-length Tau fibrillization process. Negatively-charged polysaccharide free-floating heparin (FFH) has thus far been the most commonly used Tau aggregation co-factor, although others including RNA, anionic lipids or small proteins are occasionally used (Limorenko et al., 2022, Chem Soc Rev., 2022, 51, 513-565).

The mechanisms by which anionic co-factors induce Tau aggregation are thought to include electrostatic charge neutralization of highly positively-charged regions on Tau, promoting changes in the local and global Tau polypeptide shape (Limorenko etal., 2022, supra). This allows folding of the aggregation-promoting Tau regions PHF6* and PHF6 into P-sheet-containing conformation (Figure la), leading to Tau fibrillization and assembly into higher-order Tau fibrils. In addition, structural and biophysical studies have shown high affinity of specific positively-charged Tau residues for heparin, notably the interfibrillar interface-forming residues in the core of the fibril folds (Figure 1).

To be clinically relevant, in vitro Tau aggregation systems must recapitulate some of the morphological, biochemical and structural features of Tau aggregates and fibrils found in the human pathologies. Several high-resolution structures of Tau fibrils composed of different Tau isoforms have been recently solved using cryoelectron microscopy (cryoEM; Lippens et al., 2019, Journal of Biological Chemistry 294, 9316-9325) including Tau filaments isolated from postmortem patient brain tissues. These structures combined with other experimental observations suggest that heparin or other co-factors used to induce Tau aggregation introduces several limitations that preclude biologically-relevant studies from elucidating the role of Tau aggregation in the pathogenesis of Tauopathies (Figure lb). First, recent studies demonstrated that in vitro FFH-induced Tau fibrils differ from pathological patient-derived fibrils at the biochemical (lack of PTM pattern due to use of the unmodified isoforms) and structural (amyloid core conformation) levels (Zhang et al, 2019, eLife 8, doi: 10.7554 eLife.43584). Other polyanionic co-factors, such as RNA, were also used to fibrilize Tau, and also showed marked differences to the pathological, as well as FFH-induced Tau fibrils. As is the case with FFH, the free-floating RNA is incorporated into the structure of the Tau fibril (Abskharon et al., 2022, Proceedings of the National Academy of Sciences, 119(15), p.e2119952119). In addition, the mature Tau fibrils in AD, corticobasal degeneration (CBD) and chronic traumatic encephalopathy (CTE) are composed of doublet Tau filaments with the interfilament interfaces differing between these disorders, which manifests in various twisted morphologies of the higher-order Tau fibrils. In contrast, conventional in vitro- produced Tau fibrils extensively comprise a single-filament and are structurally and morphologically divergent from any Tau fibrillar structures derived from ND patients’ brains, recently illustrated by cryoEM-solved FFH-induced fibril structures (Zhang et al., 2019, eLife 8, doi: 10. 7554/eLife.43584). Second, the adding heparin to Tau results in heterogeneous fibril populations, whereas the ultrastructures of pathological Tau fibrils are consistent between patients and are a signatory of the specific Tauopathies (Figure lb). Third, heparin and other co-factors have been shown to bind strongly and remain associated with the Tau fibril, which could interfere with their interaction with other small molecules and ligands (e.g., RNAs, small-molecule compounds) (Sibille et al., 2006, Biochemistry, 45, 12560-12572, doi:10.1021/bi060964o; Al- Hilaly et al. 2018, Journal of molecular biology, 430, 4119-4131, doi:10.1016/j.jmb.2018.08.010). These properties therefore preclude the use of FFH-induced Tau fibrils to investigate the structural basis of Tau aggregation, cellular uptake, and toxicity or their use for the development and testing of Tau-targeting and binding molecules, as well as positron emission tomography (PET) tracers.

To avoid the use of co-factors, one can use human brain-seeded Tau but no evidence has yet been shown to suggest that the structures of the patient brain-seeded full-length Tau in vitro aggregates recapitulate the pathological folds (Amyloid Atlas for available structures https://people.mbi.ucla.edu/sawaya/amyloidatlas/). More importantly, the availability of the human brain-derived material to serve as aggregation seeds is such a tremendous bottleneck in the field due to obvious issues of obtaining human post-mortem tissue from mostly elderly patients who suffered for decades from extensive dementia, that it is both impractical and ethically-desirable to be substituted with more accessible in vz/ro-produced material. As a result, for decades, investigating Tau fibrillization in vitro has required the addition of polyanions or other co-factors to induce its misfolding and aggregation, with heparin being the most commonly used. However, heparin-induced Tau fibrils exhibit high morphological heterogeneity and a striking structural divergence from Tau fibrils isolated from Tauopathies patients’ brains at ultra- and macro-structural levels.

Therefore, there is a need for the development of cost and time effective methods of preparing clinically relevant Tau fibrils which would be useful for investigating the pathophysiology of disease-relevant Tau aggregates, the development of Tau pathology -targeting and modifying therapies and PET tracers that can distinguish between different Tauopathies.

Summary of the Invention

A general object of this invention is to provide a method for the preparation of aggregates of amyloidogenic proteins that need a co-factor to aggregate (for example, prion, TDP-43), in particular Tau aggregates which are structurally more closely resembling the naturally occurring aggregates, e.g. brain-derived Tau fibrils. This invention provides a platform to achieve the desired pathological protein polymorphs through the addition of supplementary cofactors and changes in buffer conditions.

One of the specific objects of this invention is to provide a method for the preparation of Tau aggregates and isolated Tau aggregates resulting thereof.

It is advantageous to provide a method of preparation of Tau aggregates which is rapid and cost efficient.

It is advantageous to provide a method of preparation of Tau aggregates from Tau monomers with a high yield (high level of interconversion of Tau monomers to aggregates).

It is advantageous to provide a method of preparation of Tau aggregates with a high degree of purity for the further use of the aggregates in biochemical and biophysical assays.

Another of the specific objects of this invention is to provide a method for the preparation of Tau aggregates which allows the preparation of co-factor-free Tau aggregates.

Objects of this invention have been achieved by providing a method for the preparation of Tau aggregates according to claim 1.

Another of the specific objects of this invention is to provide Tau aggregates that are morphologically closely resembling to brain-derived Tau aggregates and which are suitable for screening assay for anti-aggregation compounds. Another of the specific objects of this invention is to provide Tau aggregates useful in the characterization Tau aggregates relevant for a range of Tau-associated pathologies.

It is advantageous to provide highly pure Tau aggregates which are free from contaminants which could interfere with the effects of anti-aggregation compounds.

Another of the specific objects of this invention is to provide Tau aggregates that are morphologically closely resembling the brain-derived Tau aggregates in sufficient quantity and with sufficient reproducibility to allow their use in assays and large drug screening campaigns to identify modulators of tau aggregation, study their toxicity and the cell-to-cell propagation mechanisms.

Another of the specific objects of this invention is to method for the identification of modulators of Tau protein aggregation which would be cost-effective and reproducible and relevant to m-vivo activity.

Objects of this invention have been achieved by providing a method according to claim 14.

Objects of this invention have been achieved by providing isolated co-factor free Tau aggregates according to claim 11.

Another of the specific objects of this invention is to provide a method using highly pure Tau aggregates for screening assay for anti -aggregation compounds.

Objects of this invention have been achieved by providing a method according to claim 15.

Another of the specific objects of this invention is to provide a method/kit useful in an assay for screening anti -aggregation compounds.

Disclosed herein is a method for the preparation of Tau aggregates comprising the steps of a) Providing a Tau protein in monomeric state in suspension in aqueous medium; b) Providing a surface with a co-factor (e g. heparin) covalently immobilized on-to this surface; c) Contacting said Tau protein monomer in an aqueous medium to trigger the formation of Tau aggregates; d) Isolating the formed Tau aggregates.

Also disclosed herein is an isolated co-factor free Tau protein aggregate (e.g. Tau fibril), wherein said co-factor free Tau aggregate is free from material extracted from a mammal.

Also disclosed herein is a method for the identification of modulators of Tau protein aggregation, wherein said method comprises a method for the preparation of Tau aggregates according to the invention, wherein a candidate compound is provided in the aqueous medium where the Tau protein monomer is present and further comprising the steps of:

Assessing Tau aggregate formation extent in presence of said candidate compound compared to Tau aggregate formation in absence of said candidate compound;

Selecting the candidate compounds based on their ability to modulate aggregation at the early stage of the aggregation process.

Also disclosed herein is a use of isolated co-factor free Tau protein aggregate according to the invention to seed (i.e. induce by the exogeneous addition) Tau aggregation in vitro and in biosensor cells.

Also disclosed herein is another method for the identification of modulators of Tau protein aggregate stability, said method comprising the steps of:

- Providing isolated co-factor free Tau aggregates;

Contacting said isolated co-factor free Tau aggregates with a candidate compound;

Assessing Tau aggregate disaggregation extent in presence of said candidate compound compared to Tau aggregate persistence in the absence of said candidate compound;

Selecting the candidate compounds based on their ability to modulate Tau aggregation at the late stage of the aggregation process.

Also disclosed herein is a kit for assaying early stage of Tau aggregation, said kit comprising at least one reaction vessel with its reaction surface pre-coated with a co-factor wherein said co-factor is covalently immobilized on-to this surface, at least one Tau protein or a mixture of Tau proteins in monomeric state in suspension in aqueous medium or in lyophilized form and instructions for use.

Also disclosed herein is a kit for assaying late stages of Tau aggregate stability, said kit comprising at least one isolated co-factor free Tau protein aggregate according to the invention in aqueous medium and instructions for use.

Also disclosed herein is the use of a kit according to the invention in a method according to the invention.

The present invention is based on the unexpected finding that a method according to the invention allows to generate co-factor-free Tau aggregates, in particular fibrils with amyloid-like features and morphological properties closely resembling to the brain-derived Tau aggregate such as fibrils. The method allows for the efficient production of a large amount of homogeneously-aggregated as assessed using biochemical and biophysical methods, cryo-EM microscopy and seeding competent heparin-free aggregates, in particular fibrils that are morphologically distinct from those produced in the presence of FFH.

The isolated co-factor free Tau aggregates according to the invention further allows the preparation of a range of pathology -relevant Tau protein aggregates. Indeed, the biochemical aggregation conditions can be adapted to lead to the formation of desired aggregate morphologies, and these aggregates are then fully characterized and verified in downstream tests to accurately represent the pathological structures found in brain.

Other features and advantages of the invention will be apparent from the claims, detailed description, and figures.

Brief Description of the drawings

Figure 1 describes the main steps of the mechanisms of heparin-induced Tau aggregation (a) while comparing the features of pathological fibrils (c, left) with those of standard in vitro fibrils obtained by standard methods (b & c, right) and of the fibrils obtained by a method of the invention “ClearTau” (c, left), a: Tau exists as an unstructured monomer in solution and is not prone to aggregation without co-factors. The positive charge density around repeat regions (amino acids 275-305, R2; 306-336, R3) prevents molecule folding and self-association. Heparin provides negative charges to compensate for the positive charge density on Tau and promote Tau folding into P-sheet-containing conformation conducive to fibrillization. b: General features of the pathology-derived Tau fibrils and conventional free-floating heparin (FFH)-induced Tau fibrils, c: Schematic representation of the induction of fibrils in a method of the invention (“ClearTau”) (left) compared to conventional FFH-induced Tau fibrillization (right). ClearTau show double-stranded fibrils, no integration of co-factor into fibrils, homogeneous population. FFH fibrils show mostly single-stranded fibrils, co-factor integrated into fibrils, heterogeneous population.

Figure l is a schematic representation of the main steps of a method according to the invention.

Figure 3 shows the characterization of 4R2N Tau fibrils obtained by a method according to the invention (ClearTau). a: ThS aggregation assay followed two independent Tau aggregation reactions, compared to the normal Eppendorf tube without additional co-factors: ThS fluorescence signal is rapidly generated only for Tau fibrils of the invention over the course of 48 h and follows a sigmoidal curve reaching the plateau phase. All reactions comprised 10 pM of Tau monomer at the inception. Measurements were performed in triplicates, b: Circular dichroism (CD) spectra of ClearTau fibrils c Electron microscopy imaging reveals the formation of the fibrils by 4R2N Tau in ClearTau aggregation reaction, with shorter fibrils appearing at 6 and 18 h, to long fibrils at day 1 and 2 of aggregation. Scale bars = 100 nm. Figure 4 shows a wider profile of fibril morphologies on an electron micrograph gallery of 4R2N Tau fibril aggregation overtime by a method according to the invention as described in Example 1. Figure 5 provides a comparison of 4R2N Tau fibrils generated by a method of the invention versus FFH methods as described in Example 2. A: ThS aggregation kinetics comparison of ClearTau and FFH-induced Tau fibrils. FFH was added to the 4R2N Tau monomer at 1 :4 molar concentration both reactions comprised 100 pM of Tau monomers. Reactions were set up as four independent replicates, and measurements were performed in triplicates, b: Width quantification of fibrils shows significantly narrower FFH fibrils. N of ClearTau = 618, N of FFH = 292 fibrils quantified, p **** «0.001. c: Electron microscopy imaging reveals the formation of the fibrillar structures in both reactions over time of differing morphologies. Scale bars = 500 and 100 nm. High magnification images reveal a clear doublet of the ClearTau fibrils. Scale bars = 25 nm.

Figure 6 in relation with Fig 5 shows a gallery of electron micrograph illustrating Tau variants’ fibrillization in both systems. The size bar = 100 nm.

Figure 7 shows seeding potency of Tau preformed fibrils with a method of the invention (“ClearTau”) as described in Example 3. a: In vitro seeding of Tau 4R2N monomer with the “ClearTau” preformed fibrillar seed at 1:4 molar ratio (identical to heparin-to-Tau ratio). Data were collected in five independent experiments in triplicates, each for ClearTau seed and FFH, and four independent experiments in triplicates for monomer-only conditions, b: FRET flow cytometry assessment of the HEK293T TauRDP301S biosensor cell line with added ClearTau 4R2N preformed fibrillar seed. IFD = integrated FRET density. Experiments were performed in triplicates, a minimum of 100,000 cells per run were sorted, c: Confocal imaging ClearTau 4R2N fibrillar seed assessment of cytoplasmic (arrows) and nuclear reporter foci (arrowheads) formation in HEK293T TauRD P301S biosensor cell line. Scale bars = 50 pm.

Figure 8 reports the characterization of Tau fibrils generated by a method of the invention from all six Tau isoforms as described in Example 4. a: Reaction endpoint at ThS fluorescence shows the signal for all isoforms, with the highest ThS signal found for ClearTau 3R2N isoforms, likely due to the unique straight fibril morphology. Three independent ClearTau aggregation reactions for each Tau isoform were set up at 100 pM for 48 h at 37 °C under shaking conditions, b: Quantification of Tau isoforms’ fibrillar widths. Tukey’s test with multiple comparisons shows significance levels at p * <0.05, **** «0.001. c: Tau isoform incorporation into the pellet fraction. Average and st. dev. of three independent reactions, d: Transmission electron microscopy assessment of ClearTau fibrils composed of different Tau isoforms. Close-ups of the boxed areas are in black squares, e: CD spectra of the endpoint of the fibrillization of all six Tau isoforms reveal the adoption of higher-order molecular structures compared to the starting isoforms’ monomers, f High magnification of the different isoform fibril morphologies, g: SDS-PAGE gel visualization of Tau protein retainment in the supernatant (S) and incorporation into insoluble pellet (P).

Figure 9 reports the characterization of monomer incorporation into the fibril-containing Tau pellet fraction across three independent repeats for each Tau isoform as described in Example 4. Three independent aggregation reactions for each Tau isoform (18 in total) were set up at 100 pM for 48 h at 37°C under shaking conditions. The ~500 ul mixtures were ultracentrifuged at 100 ’000g for 1 h, supernatant was removed, pellets washed in dH2O twice and resuspended in 500 pl of dH2O. PAAG gels were run by loading 10 pl of supernatant or pellet + 10 pl 2X Laemmli buffer, or 5 pl Whole + 15 pl Laemmli buffer. Whole samples were loaded at half the amount to prevent the oversaturation. Whole sample signals were not included in the quantification, a: Workflow of the fractionation protocol to separate the Tau fibrils (pellet) and monomers (supernatant), b: Quantification of the monomer-to-fibril proportion of all 18 reactions shows efficient incorporation of monomers into the fibrillar fraction as early as at 48 h of the reaction in the current set up. c: The whole, supernatant and pellet fractions were loaded on the SDS-PAGE gel and stained with the Coomassie total protein stain to assess fibrillization efficiency and monomer incorporation. All isoforms across all three independent repeats show efficient formation of the fibrils separated into the pellet fractions.

Figure 10 shows cryo-EM micrographs of fibrils from ClearTau 4R2N (a) and ClearTau 3R2N (b) proteins as described in Example 4. Selected singlets and doublets are outlined with red or yellow boxes. Representative 2D class averages of singlets and doublets from large 900-pixel segments downscaled to 300 pixels with a visible twist of the amyloid core. Representative 2D class averages of singlets from 300-pixel non-scaled segments with clear amyloid core and 4.77 A separation of beta-strands. Scale bars = 10 and 50 nm. (c) TEM images demonstrate the doublet composition of ClearTau fibrils of isoforms 4R2N and 3R2N (arrows). 4R2N protein fibrils were proteolytically digested by protease K (PK) to reveal the unwinding of the singlet fibrils composing the higher- order doublets.

Figure 11 presents the characterization of Tau fibrils generated from Tau isoform mixtures by a method of the invention as described in Example 5. a: Three independent reactions were set up to include equal amounts of indicated Tau isoforms for 48 h for each condition. The fibrils (pellet fraction) were isolated from the remaining monomers (supernatant fraction) by ultracentrifugation. On the right: expected banding patterns of SDS-PAGE gel Tau isoform under four reaction conditions. (Figure 6). b: Quantification of the percentage of the monomer incorporation into the fibrils for all reactions shows high efficiency as early as 48 h of the aggregation, c: Representation of the quantification of the relative isoform incorporation into the fibrils for all repeats of the four mixtures, d: ThS fluorescence of the endpoint fibrils for all reactions, e: Electron microscopy assessment of ClearTau fibrils prepared from the mixtures of Tau isoforms. All samples contain well-defined doublet fibrils, except the All isoforms (Sample 2), which appears to comprise the straight, rigid singlet filaments. Scale bars= 200 nm. f: Quantifications of the fibril widths for each repeat of the four mixtures. Tukey’s test with multiple comparisons shows significance levels at * <0.05, ***<0.01, **** <0.001. SDS-PAGE gels were run by loading 10 pl of supernatant or pellet + 10 pl 2x Laemmli buffer, or 5 pl Whole + 15 pl Laemmli buffer. Whole samples were loaded at half the amount to prevent oversaturation. Whole sample signals were not included in the quantification, g: Correlation analysis of fibril with ThS, monomer incorporation into the pellet and ThS fluorescence.

Figure 12 relating to Figure 11 presents annotated SDS-PAGE gel of the Tau isoform mixtures shows relative incorporation of the Tau isoform monomers into the Tau fibrils of the invention as described in Example 5. The circle size besides each band indicates the relative amount for better perception. The whole samples contain equal amounts of all isoforms as expected. Across the triplicate repeats for all mixtures, the 4R-containing Tau isoforms were more efficiently incorporated into the fibrils. The notable exception was All isoforms Sample 2 (grey thick line box), that demonstrated the complete incorporation of all the Tau monomers into the fibrils, at the same time showing the highest thioflavin S fluorescence and singlet Tau filaments. The reactions were performed at 100 pM initial Tau concentration for 48 h with orbital shaking at 1000 rpm at 37°C. The -500 pl mixtures were ultracentrifuged at lOO’OOOg for 1 h, supernatant was removed, pellets washed in dFhO twice and resuspended in 500 pl of dFLO. Polyacrylamide gels were run by loading 10 pl of supernatant or pellet + 10 pl 2X Laemmli buffer, or 5 pl Whole + 15 pl Laemmli buffer. Whole samples were loaded at half the amount to prevent the oversaturation.

Figure 13 shows the ClearTau fibrillization in the presence of co-factors as described in Example 6. a: Electron microscopy imaging; b: SDS-PAGE gel of Tau fibrils in the presence of polyU RNA. c: Electron microscopy imaging; d: SDS-PAGE gel of Tau fibrils in the presence of adenosine triphosphate (ATP). TO - time 0, W - whole sample, S - soluble fraction, P - pellet fraction. Both reactions were performed in triplicate independent aggregation reactions.

Figure 14 illustrates the use of the method of the invention for fibrillization of full-length Tau 4R2N with mutation P301L as described in Example 7. a: Schematic of proteinase K limited proteolysis to remove the “brush” residues to reveal the fibril core structure, b: EM of ClearTau and FFH-fibrillized Tau 4R2N P301L. Scale bars = 100 nm. c: Proteolytic digestion of the fibrils by protease K (PK). Arrows point out the twists on the ClearTau fibrils, absent in the FFH samples. Scale bars = 50 nm. d: Quantification of the widths of fibrils pre- and post-PK digestion, p **** «0.001. e: Representative gels demonstrating the monomer incorporation at 24h aggregation reaction for Tau 4R2N P301L ClearTau and FFH systems (complete gels and individual replicates Figure 10a, Figure 15a). f: Schematic of RNA-binding assay workflow, g: PolyU RNA-binding capacity of the 4R2N P301L fibrils prepared using ClearTau or FFH methods was assessed in the aggregation buffer, or in the higher ionic strength aggregation buffer (+100 mM MgCb).

Figure 15 relating to Figure 14 shows monomer incorporation into the fibrils and RNA-binding capacity of Tau 4R2N P301L in the claimed method and in the presence of FFH. A: Three independent samples were fibrillized at 100 pM for 24h, the -500 pl mixtures were ultracentrifuged at 100 ’000g for 1 h, supernatant was removed, pellets washed in dH2O twice and resuspended in 500 pl of dH2O. Polyacrylamide gels were run by loading 10 pl of supernatant or pellet + 10 pl 2X Laemmli buffer. B: PolyA and tRNA binding capacity of 4R2N P301L fibrils in the aggregation buffer and in the higher ionic strength buffer (+ 100 mM MgCh).

Figure 16 shows FRET flow cytometry bivariate plots. FRET vs. CFP donor bivariate plots depict the FRET-negative HEK293T TauRDP301S biosensor cells used to configure the gating strategy to detect FRET -positive ClearTau-seeded HEK293T TauRDP301S biosensor cells as described in Holmes, 2014, PNAS 111.41, E4376-E4385.

Figure 17 illustrates the implementation of a method and assay according to the invention as described in Example 8 for the identification of agents with aggregation-modifying properties, a: Schematic illustration of an assay (“ClearTau”) workflow, b: SDS-PAGE illustrating the distribution of the Tau in the soluble (S) and pellet (P) fractions after incubation in the presence of candidate molecules, c: Quantification of the molar ratios of Tau distribution into soluble or pellet fraction after aggregation reactions. Chevron direction indicates increase or decrease of Tau aggregation in the presence of candidate molecule relative to the reaction in absence of candidate molecule, d: Quantifications of the fibril aggregate widths for each repeat of the four mixtures. ANOVA tests with multiple comparisons shows significance levels at * <0.05, ***<0.01, **** <0.001. e: Electron microscopy assessment of fibrils of the invention and oligomers (arrowheads) formed in the presence of a candidate molecule.

Figure 18 illustrates variety of the applications of the methods and the resulting aggregates according to the invention (“ClearTau system”) for production and study of Tau aggregates as described in Example 9. a: Aggregation of 4R2N at lower shaking speed (300 rpm at 37 °C) results in more tightly-wound fibrils, b: Aggregation of 4R2N Tau in the presence of mouse cortical or HEK293T cell line lysate extracts, c: aggregation and nitration of wild-type Tau isoforms 3R2N, 4R0N, 4R1N, 4R2N, as well as 4R2N isoform mutants, containing a single tyrosine at Y18 (4F/Y18) or Y29 (4F/Y29); two tyrosines at both Y18 and Y29 (3F/Y18/Y29); three tyrosines at Y197/Y310/Y394 with phenylalanines at residues 18 and 29 (Y18F/Y29F); and four tyrosines with either Y18 (Y18F) or Y29 (Y29F) mutated to phenylalanine.

Figure 19 supports the re-usability of the covalently-coated aggregation vessels (here Eppendorf tubes) for multiple rounds of the aggregation reactions of 3R2N and 4R2N/Y18F Tau isoforms as described in Example 10. a: Schematic of sequential washes of the aggregation tubes twice with PBS with vortexing at 1 min, followed by twice with 1% Hellmanex detergent with vortexing at 1 min, followed by thrice with PBS with vortexing at 1 min and drying, b: SDS-PAGE gels and TEM images of the aggregated Tau fibrils after up to five tube washes. The gels demonstrate the shift of protein into insoluble fractions (arrows) for all isoforms and all washes, and EM images further demonstrate Tau fibrils at each sequential aggregation.

Figure 20 provides a schematic depiction of the possibilities offered by a method of the invention and of the obtained Tau aggregates (such as fibrils or oligomers) as a platform to accelerate reconstruction of pathology-resembling Tau aggregates in vitro for research into Tau aggregation processes and development of Tau fibril-targeting therapies and imaging agents. New insights about various co-factor molecules linked to Tau pathology in the brain, or those that induce Tau pathology in a controlled in vitro environment, will help determine the choice of co-factors like polysaccharides, nucleotides, lipids, proteins, and more. These co-factors can be anchored to the surface of the reaction vessel, which enables them to trigger and facilitate Tau polymerization while ensuring they don't become part of the growing Tau fibrils. The composition of Tau fibrillar aggregates can be tailored to contain specific Tau isoforms, post-translational modifications and other molecules, to more faithfully represent the pathological aggregates. The lack of free-floating co-factors prevents undesirable reactivity with the components in the aggregation reaction. Fine- tuning reaction conditions, such as temperature, agitation, or buffer, helps our ability to produce in vitro Tau fibrils more closely resembling the pathology-derived Tau aggregates. Method of the invention can be applied to target the early stages of Tau aggregation for screening of aggregationmodifying compounds. The method and co-factor-free aggregates of the invention can be used for research into Tau aggregation mechanisms, Tau seeding and spreading in in vitro and in vivo models, as well as to develop Tau fibril-targeting antibodies or positron emission tomography tracer molecules.

Detailed description of embodiments of the invention

The expression “Tau fibril” refers to high-molecular weight Tau aggregate containing cross- sheet core structure, polymerically fibrillized by addition of monomers to the growing end of fibrils. The expression “ClearTau™ fibrils” refers to Tau fibrils according to the invention. ClearTau™ fibrils can be obtained by a method according to the invention (“ClearTau™ method or ClearTau™ system”).

The expression “Tau seed” refers to sonicated Tau fibrils.

The expression “co-factor free Tau protein aggregate or fibril” refers to ClearTau™.

The expression “Tau protein” refers to full-length or truncated Tau proteins and all of its isoforms that could be post-translationally modified, chemically modified, synthetic or semisynthetic, in combination with other cofactors, and mixtures thereof.

The expression “surface” refers to any support on-to which a co-factor could be covalently attached to and on-to which an aqueous suspension of Tau protein in monomeric state can be contacted with in order to carry out an aggregation reaction. It includes a plate, a wall or a plane of a reaction vessel. This can include Eppendorf tubes, 96 or 384 well plates.

The expression “candidate modulator” refers to an agent of interest for its possible impact on Tau aggregation (e.g. small molecules, antibodies etc.), either modulating the formation of Tau aggregates or inducing disaggregation of already formed Tau aggregates. Candidate modulators include aggregation inhibitor and disaggregating candidate compounds.

The expression “aggregate” includes the formation of a multimeric assembly of the protein which includes fibrils, fibrillar species, soluble and insoluble oligomeric species such as described in Limorenko et al., 2022, Chem. Soc. Rev., 51, 513-565.

The expression “co-factor” refers to agents that are needed to trigger the aggregation of the amyloidogenic protein, and/or sustain polymerization of the protein into a conformer of consistent ultrastructure, morphology and biochemical properties. Those include anionic cofactors and polysaccharides such as heparin, sulfated or modified heparin, heparan sulfate, keratan sulfate, dermatan sulfate and chondroitin sulfates such as chondroitin-4-sulfate and chondroitin-6-sulfate and other related glycosaminoglycans (GAGs) such as dextran sulfate, as well as negatively- charged nucleic acid, lipid molecules, metals and metal complexes, and synthetic and natural compounds and metabolites. Heparin and other polysaccharide coating can be achieved by known methods such as described in Linhardt, 2008, Curr Top Med Chem 8, 80-100. Other immobilized co-factors include nucleic acids such as RNA and lipids.

The expression “co-factor free” refers to the absence of co-factor in the material. The material obtainable according to a method of the invention is a co-factor free material, even in absence of purification. Referring to the figures, in particular first to Figure 2, is provided an illustration of a method for the preparation of Tau aggregates, in particular fibrils according to an embodiment of the invention. The illustrated method generally comprises the steps of: a) Providing a Tau protein in monomeric state in suspension in aqueous medium; b) Providing a surface with a co-factor (e g. heparin) covalently immobilized on-to this surface; c) Contacting said Tau protein monomer in an aqueous medium with said surface; d) Isolating the resulting Tau aggregates, in particular fibrils.

According to a particular aspect, Tau protein in monomeric state is provided at a concentration from about 5 pM to about 100 pM (e.g. 5 to about 10 pM) in suspension in aqueous medium.

Depending on the applications, the concentration of the Tau protein in monomeric state will be adapted. For example, for functional assays a concentration of about 10 pM is generally used, and for bulk production of fibrils of about 100 pM.

According to a particular aspect, the Tau protein in monomeric state is a full-length Tau monomer.

According to a particular aspect, Tau protein in monomeric state is in suspension in aqueous medium, wherein said aqueous medium is phosphate-buffered saline (PBS).

According to a particular aspect, Tau protein in monomeric state is in suspension in aqueous medium at a pH of about 7.4.

According to a particular aspect, Tau protein in monomeric state is contacted with a surface with a co-factor (e.g. heparin) covalently immobilized on-to this surface for 24 to about 48 hours before isolating the Tau aggregates, in particular fibrils.

According to a particular aspect, Tau protein in monomeric state is contacted with a surface with a co-factor (e.g. heparin) covalently immobilized on-to this surface at a reacting temperature from about 22 to about 37°C.

According to a particular aspect, Tau protein in monomeric state is contacted with a surface with a co-factor (e.g. heparin) covalently immobilized on-to this surface under orbital shaking of said surface (e.g. at 100-1000 rpm).

According to a particular aspect, Tau protein in monomeric state is contacted with a surface with a co-factor (e.g. heparin) covalently immobilized on-to this surface wherein said surface is a wall of a reaction vessel or of muti-well plates. According to a particular aspect, the co-factor (e.g. heparin) molecules are covalently immobilized on-to on at least one of the walls of a reaction vessel.

According to a particular aspect, the co-factor (e.g. heparin) molecules are covalently immobilized by amide coupling.

According to a particular aspect, the co-factor is heparin. In particular, the immobilized co-factor is non-sulfated 5 kDa immobilized heparin.

According to a particular aspect, the co-factor is covalently immobilized by coupling an acid group to the free amine group on the functionalized vessel surface. The maximum coupling density will be determined by a maximum saturation allowed by the available free amine groups and steric hindrance of the polysaccharide molecules that depend on the size and modifications thereof.

According to a particular aspect, Tau-to-cofactor molar ratio can be adjusted for achieving optimized aggregation at a desired Tau concentration.

According to a particular aspect, typically, the Tau protein in monomeric state is provided at a molar ratio to the cofactor, in particular heparin from about 1 :4 to about 1 :2.

According to a particular aspect, Tau aggregates, in particular fibrils are isolated by ultracentrifugation and resuspension of aggregates, in aqueous medium (e.g. dITO or another buffer).

According to a particular aspect, is provided an isolated co-factor free Tau fibril obtainable from a method according to the invention.

According to a particular aspect, is provided an isolated co-factor free Tau fibril obtained by a method of the invention and which is structurally closely resembling to naturally occurring aggregates, in particular fibrils, and serves as a useful tool to achieve desired pathological polymorphs through the addition of supplementary cofactors and changes in buffer conditions. Cofactor free Tau fibril can be characterized for their structural similarity with naturally occurring aggregates by micro- and nano-meter scale imaging (i.e. electron, atomic force, cryo-electron microscopy) on the morphology, structure and ultrastructure characteristics, including but not limited to aggregate/fibril/oligomer shape, size, length, curvature, straightness, twisting handedness, twist spacing, number of protofilaments composing the fibril, Young’s modulus, brittleness, fragmentation etc.

According to a particular aspect, the method according to the invention allows to obtain high yields (e.g. up to about 100%) co-factor-free Tau protein fibrils. According to a particular aspect, isolated co-factor free Tau protein fibril according to the invention can be used to seed (i.e. induce by the exogeneous addition) Tau aggregation in vitro and in biosensor cells.

According to a further particular aspect, is provided a method to seed Tau aggregation in vitro, said method comprising contacting isolated co-factor free Tau protein fibril according to the invention (seeds) with cell lines stably expressing tagged Tau fragments under transduction conditions, for example as described in Holmes et al., 2014, Proceedings of the National Academy of sciences, 111(41), pp.E4376-E4385. According to a particular aspect, the formation of aggregates, in particular fibrils can be detected through the detection of a formed fluorescent foci using microscopy or flow cytometry.

According to another further particular aspect is provided a method for the identification of modulators of Tau protein aggregation, wherein compounds able to induce disaggregation of formed fibrils are identified.

According to another further particular aspect is provided a method for the identification of modulators of Tau protein aggregation, according to the invention, wherein Tau aggregate, in particular fibril formation extent is assessed by electron microscopy.

According to another further particular aspect is provided a method for the identification of modulators of Tau protein aggregation, wherein inhibitors of the early stages of the Tau aggregation process are identified.

According to another further particular aspect is provided a method for the identification of modulators of Tau protein aggregation, said method comprising the following steps:

- Providing a Tau protein in monomeric state in suspension in aqueous medium;

- Providing a surface such as 96 or 384 well plates with a co-factor (e.g. heparin) covalently immobilized on-to this surface;

Contacting said Tau protein monomer in an aqueous medium with said surface in presence of a candidate modulator (e.g. aggregation inhibiting candidate) to the aqueous medium;

Assessing Tau aggregate, in particular fibril formation extent in presence of said candidate compound compared to Tau aggregate formation in absence of said candidate compound;

Selecting the candidate compounds based on their ability to inhibit the aggregation process at the early stage.

According to a particular embodiment, the Tau protein is contacted a concentration from about 5 pM to about 100 pM (e.g. 5-10 pM) for about 24 to about 48 hours. According to a particular embodiment, the Tau protein monomer in an aqueous medium is preincubated with compound before contacting the surface with a co-factor (e g. heparin) covalently immobilized on-to this surface.

According to another further particular aspect is provided a method for the identification of modulators of Tau protein aggregation, said method comprising the following steps:

- Providing an isolated co-factor free Tau protein aggregate, in particular fibril according to the invention in suspension in aqueous medium;

Contacting said co-factor free Tau protein aggregate, in particular fibril, in aqueous medium with a candidate modulator (e.g. disaggregation candidate) and incubating the mixture for at least 24h;

Assessing Tau aggregate, in particular fibril formation extent in presence of said candidate modulator compared to Tau aggregate formation in absence of said candidate modulator;

Selecting the candidate modulators based on their ability to destabilize the aggregation process at the late stage.

According to another particular embodiment, the Tau aggregate formation extent can be assessed by using amyloid-binding dyes (Congo Red, thioflavin S (ThS), Amytracker) indicating fibril formation with signal enhancement (ThS) and electron microscopy for detecting the fibril morphology.

According to another particular embodiment, the ability of a candidate compound to disaggregate Tau aggregates, in particular fibril can be followed by using amyloid-binding dyes (Congo Red, thioflavin S, Amytracker) and following the signal dissipation indicating the loss of aggregatedd structures.

The lack of loose heparin contamination would also allow for the ultrastructural assessment of the Tau-candidate compound formed complexes.

Also disclosed herein is a kit for assaying early stage of Tau aggregation, said kit comprising at least one reaction vessel with its reaction surface pre-coated with a co-factor wherein said co-factor is covalently immobilized on-to this surface, at least one Tau protein or a mixture of Tau proteins in monomeric state in suspension in aqueous medium or in lyophilized form and instructions for use and optionally further comprising at least one of the following elements:

- an aggregation buffer solution such as for example phosphate buffer (e g.10 mM Phosphate, 50 mM NaF, 0.5 mM fresh DTT);

- an amyloid-detecting fluorescent probe (e.g. ThS, ThT, CongoRed etc.). According to another particular aspect is provided a method for the identification of modulators of Tau protein aggregation according to the invention wherein said method comprises the steps of:

Providing a kit for assaying early stage of Tau aggregation according to the invention;

- Mixing together the reaction components of the kit (Tau monomers, aggregation buffer and candidate modulator) into at least one reaction vessel with its reaction surface pre-coated with a co-factor covalently immobilized on-to this surface;

Incubating the obtained reaction mixture for at least 24h;

Supplementing the reaction mixture with at least one amyloid-detecting fluorescent probe; Detecting the extent of Tau aggregate, in particular fibril formation using a fluorescence reader;

Comparing the extent of Tau aggregate, in particular fibril formation with a positive control (e g. in absence of the candidate modulator), wherein a decrease of fluorescence compared to a positive control is indicative of the ability of a candidate modulator in preventing the aggregation of Tau (i.e. at least a fraction of the Tau protein remains in soluble form without formation of aggregate, in particular fibril).

According to a particular embodiment, a negative control sample comprising the aggregation buffer without the Tau monomers but including the candidate modulator is used control for candidate modulator autofluorescence and fluorescence intensities are standardized to this value.

A kit according to the invention can be further used in combination with downstream biophysical or biochemical assays such as for example electron and atomic force microscopy and circular dichroism for the further characterization of the candidate modulator or aggregates.

Also disclosed herein is a kit for assaying late stages of Tau aggregate, in particular fibril stability, said kit comprising at least one isolated co-factor free Tau protein aggregate, in particular fibril according to the invention in aqueous medium and instructions for use and optionally further comprising at least one of the following elements: at least one amyloid-detecting fluorescent probe (e.g.ThS, ThT, CongoRed etc.); at least one reaction buffer such as phosphate buffer (e.g.10 mM Phosphate, 50 mM NaF, 0.5 mM fresh DTT).

According to another particular aspect is provided a method for the identification of modulators of late stages of Tau aggregate, in particular fibril stability according to the invention wherein said method comprises the steps of:

- Providing a kit for assaying late stages of Tau aggregate stability according to the invention; - Mixing together the reaction components of the kit (least one isolated co-factor free Tau protein aggregate, reaction buffer and candidate modulator) into at least one reaction vessel; Incubating the obtained reaction mixture for at least 24h;

Supplementing the reaction mixture with at least one amyloid-detecting fluorescent probe;

- Detecting the extent of Tau aggregate persistence using a fluorescence reader;

Comparing the extent of Tau aggregate persistence with a positive control (e.g. intact Tau aggregates in absence of the candidate modulator), wherein a decrease of fluorescence compared to a positive control is indicative of the ability of a candidate modulator in remodeling/disaggregating the aggregates into non-amyloid containing species, such as monomers.

According to a particular embodiment, a negative control sample comprising only the reaction buffer without the Tau aggregates but including the candidate modulator is used control for candidate modulator autofluorescence and fluorescence intensities are standardized to this value.

According to a particular embodiment, the kit allows to investigate late events in the Tau aggregation cascade, such as stability of the aggregates and their seeding potential to induce formation of new aggregates, for screening of agents preventing these events. Those kits can be further used in combination with downstream biophysical or biochemical assays for the characterization of aggregate morphology, stability and biochemical qualities for example by imaging and seeding assays.

According to one aspect, the high interconversion levels of Tau monomer to aggregate, in particular fibrils in the present method are fully compatible the use of high-cost, ultrapure, post-translationally modified, chemically modified, synthetic or semisynthetic protein material since it prevents the waste and loss of the material.

According to another aspect, the lack of association of the co-factors with the Tau protein aggregate, in particular fibrils allows downstream use in the ultrasensitive biochemical and biophysical assays without artefacts, in particular for the development of antibodies targeting Tau fibrillar structures or the identification of molecules for their effects on the Tau aggregation. The present method allows to produce Tau aggregate, in particular fibrils in absence of intercalation and binding of the heparin to the growing Tau aggregate, in particular fibrils, which has been never achieved.

According to another aspect, the Tau aggregates of the invention are structurally similar to pathological Tau aggregates, notably those are characterized by the presence of doublet strand formations from homomorphic singlet strands that associate through the inter-protofilament interface, resembling pathological Tau fibrils derived from Tauopathy patients, as revealed by cryoelectron microscopy.

Finally, the Tau fibrils of the invention present the unique advantages to retain the ability to bind RNA, share some structural/morphological features with brain-derived pathological fibrils and induce efficient seeding of Tau aggregation in vitro and in biosensor cells.

According to another aspect, the Tau aggregates of the invention can be used in ultrasensitive biochemical and biophysical assays (e.g. for identifying disaggregation molecules), as well as in cell and in vivo studies in cell lines, primary cultures, and non-mammalian and mammalian animal models using direct treatment with fibrils by intracranial or systemic fibril delivery.

Therefore, the method, the aggregates and the uses thereof represent useful tools for use in the diagnosis and identification of agents useful for the prevention or treatment of Tauopathies.

The invention having been described, the following examples are presented by way of illustration, and not limitation.

EXAMPLES

Example 1: Method for generating heparin-free aggregate, in particular fibrils according to the invention

To determine if surface-immobilized heparin retains the ability to induce Tau fibrillization, the extent of fibril formation for the variant Tau 4R2N monomers (10 pM) in 1.5 ml Eppendorf tubes that are covalently coated by amide coupling without linker with heparin molecules and in standard 1.5 ml Eppendorf tubes containing FFH (Tau to heparin ratio at 4:1) has been assessed and compared using ThS aggregation assay, electron microscopy, seeding assays, fractionation assay, RNA-binding assay, co-aggregation with RNA and ATP co-factors while using a method of the invention and a standard FFH method where co-factor is free-floating in the solution.

Tau fibrils were prepared using a method according to the invention following the protocol below: a) Providing a Tau protein in monomeric state in suspension in aqueous medium

Monomeric human full-length 4R2N Tau was obtained as described below and was diluted to 100 pM in phosphate buffer (10 mM Phosphate, 50 mMNaF, 0.5 mM fresh DTT) to obtain a suspension of the monomers in aqueous solution (Ait-Bouziad et al. supra). b) Providing a surface with a co-factor (e.g. heparin) covalently immobilized on-to this surface

A heparin-coated reaction tube was provided (ThermoFisherScientific) where heparin was immobilized by amide coupling without linker as described in (Linhardt et al., 2008, supra). c) Contacting said Tau protein monomer in an aqueous medium with said surface The aqueous solution was then contacted to the inner surface of the heparin-coated reaction tube (ThermoFisherScientific). The reaction mixture was incubated for 24 h with orbital shaking at 1’000 rpm (Peqlab, Thriller) at 37°C. d) Isolating co-factor -free fibrils

The reaction mixture was then ultracentrifuged using Beckman-Coulter ultracentrifuge at 100’000 g for 1 h at 4°C, the supernatant was then removed and discarded, the pellet was washed twice with dFBO. The pellet containing fibrils was resuspended in dHiO to 100 mM, aliquoted to single-use aliquots and stored at -80°C.

Analogous measurement procedures were followed for other Tau isoforms 4R1N, 4R0N, 3R2N, 3R1N, 3R0N, and mixtures thereof, mutant 4R2N P301L, fragments K18 and K19.

The isolated fibrils were then analyzed by electron microscopy and ThS fluorescence as described below and compared to FFH fibril seeds.

The kinetics of aggregation was followed by the thioflavin S fluorescence (ThS) assay over the course of 48 h. As shown in Figure 3a, a rapid increase in ThS fluorescence was only observed in the heparin-coated tubes, whereas no change in ThS signal was detected in the standard Eppendorf tube without immobilized or FFH at this concentration. Circular dichroism (CD) measurement of the samples during the reaction revealed a significant drop in the CD signal, suggesting that the majority of the Tau monomer have converted to insoluble Tau fibrils that precipitate out of solution as early as 1 day into the reaction progression (Figure 3b). Electron microscopy (EM) imaging revealed the formation of short fibrillar aggregates at 6 and 18 h into the reaction time course, with the dominant appearance of long curvy fibrils at 24 h and later (Figure 3c, Figure 4).

These results demonstrate that the method of the invention is efficient for the fibrillization of Tau, is easy to implement and is highly versatile.

Example 2: Main differences between Tau fibrils according to the invention and FFH- induced Tau fibrils

The kinetics of aggregation of Tau 4R2N using the method of the invention were compared to those obtained in the conventional FFH method at a high concentration of 100 pM, which favors the formation of fibrils. As shown in Figure 5a, a rapid increase in ThS fluorescence as early as 8 h was observed using both the methods, illustrating the comparable kinetic profiles of Tau fibrillization at this concentration. EM imagining of the fibrils revealed the formation of the fibrils under both conditions. However, the fibril morphologies substantially differed (Figure 5b and c). Tau fibrils of the invention were long, wavy and homogeneous, with no intermediate structures or amorphous aggregates. In contrast, FFH-induced Tau fibrils were heterogeneous and contained amorphous structures. These results suggest that Tau fibrils of the invention could differ in their ultrastructure properties. Therefore, the fibril widths of both types of Tau fibrils were measured and quantified (Figure 5b). The mean width of the Tau fibrils obtained by the method of the invention was greater than the FFH-induced fibrils (15.2 and 13 nm, respectively, p«0.001).

The aggregation of 4R2N, 4R0N, 3R2N, 3R0N, mutant 4R2N P301L, and fragments K18 and K19 was also assessed and compared using both systems. The EM imaging revealed consistently longer and more homogeneous Tau fibrils of the invention (Figure 6) and illustrated the versatility of the ClearTau method for a variety of Tau proteins and fragments.

These results show that the method of the invention 1) works on experimentally-relevant time scales; 2) is likely devoid of the OFF-pathway amorphous aggregates; and 3) produces long homogeneous fibrils that have the potential to be reminiscent of the pathological Tau fibrils.

Example 3: Seeding potency of the Tau fibrils of the invention in vitro and in cells

To assess whether 4R2N Tau fibrils of the invention were competent to seed aggregation of the Tau monomer in the absence of any co-factors in vitro, the aggregation assay was performed using a microplate setup (Figure 7a). Tau 4R2N monomers were seeded with either sonicated 4R2N ClearTau or FFH fibrils. No increase of fluorescence was observed in ClearTau-seeded samples in the absence of free-floating cofactors in the reaction or their presence during the process of seed formation.

The formation of fluorescent foci in the HEK293T TauRD P301S biosensor (BS) cell line is commonly used to assess Tau seeding activity (Holmes et al., 2014, Proceedings of the National Academy of Sciences of the United States of America, 111, E4376-4385, doi:10.1073/pnas,1411649111; Lo, 2021, Bioeng Transl Med, 6, el0231, dot: 10.1002/btm2. 10231). These BS cells stably express the repeat domain of Tau with the P301S mutation (TauRD P301 S) fused with a yellow fluorescent protein (YFP) or cyan fluorescent protein (CFP). The FRET signal is generated upon the reporter foci formation in these cells, where the two fluorophores, YFP and CFP, come in close proximity for the wavelength energy transfer, the signal from which is then detected and quantified using flow cytometry set up as initially described (Furman & Diamond, 2017, Tau Protein, 349-359 (Springer)).

The results in Figure 7b demonstrate the FRET signal's detection contingent on the amount of the fibrillar seeds added, demonstrating the dose-dependency of the seeding with ClearTau fibrils in this cellular model. In addition, confocal imaging verified the presence of both cytoplasmic (Figure i

4c, red arrows) and nuclear reporter foci (Figure 7c, arrowheads), whereas none were present in the non-transduced cultures.

These results demonstrate that Tau seeds obtained by a method according to the invention are a valuable tool to decipher the Tau aggregation mechanism and its contingency on the co-factors in seeding Tau monomer in vitro, as well as in the induction of reporter foci in the HEK293T TauRD P301S cell line.

Example 4: Ability of the method of the invention to induce efficient fibrillization of isoforms, mutant and fragments of Tau

In human CNS, Tau is expressed as six isoforms differing in the number of incorporated N-terminal repeats, and differential inclusion of the R2 domain in the MTBR. As different isoforms contribute to Tau aggregation in the specific Tauopathies, it is important to develop high-fidelity methods for in vitro production of Tau fibrils of all six Tau isoforms or mixtures of the specific isoforms. Therefore, the applicability of the method of the invention to fibrillization of Tau 4R2N, 4R1N, 4R0N, 3R2N, 3R1N and 3R0N proteins was assessed. Three independent reactions were conducted for each of the six Tau isoforms with the method of the invention, and resulting aggregates were evaluated using ThS assay, CD and EM.

Each Tau isoform was fibrillized as described previously. The extent of fibrillization was assessed on the whole sample aliquots by ThS fluorescence assay before ultracentrifugation separating monomers and pellets (Figure 8a), showing the signal for all Tau isoforms and repeats. However, inter-isoform variability was observed with 3R2N showing the highest ThS fluorescence. Then, the fibrils were separated by ultracentrifugation to yield pure fibril fraction for further analyses. The quantification of the ClearTau fibril widths at the widest part with respect to the twists (if present) of all Tau isoforms showed the shortest isoform 3R0N having the narrowest fibrils at 12.49 ± 2.59 nm (Figure 8b, Table 1 below), followed by 3R1N at 14.49 ± 3.19 nm, 3R2N at 14.73 ± 2.55 nm, 4R2N at 15.20 ± 2.24 nm, 4R0N at 15.70 ± 3.04 nm, with widest fibrils shown by 4R1N isoform at 15.83 ± 3.49 nm.

Table 1

The extent of monomer incorporation into fibrils showed the highest efficiency for 4R0N and all 3R, as evident by the shift of Tau from soluble (S) to insoluble pellet (P) fractions isoforms (Figure 8c and g, Figure 9). Examination of the fibrils by EM reveals morphological differences between the fibrils derived from the various Tau isoforms (Figure 8d and f). The 4R2N, 4R1N and 4R0N samples formed long, flexible curved fibrils. 3R2N formed straight, more rigid fibrils. 3R1N and 3R0N showed curved, twisted fibrils. The 3R2N stood out with non-twisted straight fibrils, whereas all other isoforms’ twists differed in their periodicity and higher-order coiling, resulting in fibrils of differing conformations (Figure 8e). The CD spectra of the aggregated solutions of the isoforms 4R2N and 3R1N demonstrated a signal loss, indicating the formation of structured insoluble material, whereas isoforms 4R1N, 4R0N, 3R2N, and 3R0N demonstrated the shift towards the higher-order secondary structure-containing species (Figure 8e). Interestingly, the 3R2N showed the most prominent shift of the minimum towards 210 nm.

Next, cryo-EM was used to gain more insight into the ultrastructural properties of Tau fibrils of the invention composed from 4R2N and 3R2N Tau isoforms. Raw cryo-EM micrographs demonstrated that both fibrils from ClearTau 4R2N (Figure 10a) and 3R2N (Figure 10b) proteins were straight and fragmented. Furthermore, an inspection of raw micrographs and 2D class averages revealed the presence of two polymorphs with different widths. The singlet polymorphs comprised long and 160 A wide filaments with major and minor grooves and visible crossover represented 64% and 73% in all of the extracted segments for samples 3R2N and 4R2N, respectively. The doublet polymorphs (36% for 3R2N and 27% for 4R2N) are short and 380 A wide filaments that appeared to be composed of two copies of 180 A wide polymorphs zipped together with minor grooves. These observations suggest that the absence of heparin at the protofibril interface allowed for the formation of fibrils composed of two intertwined filaments - the doublets. Interestingly, the edges of the protofibrils exhibited different lengths and lacked a uniform flat edge. This suggests the formation of doublets is preceded by the formation of singlet fibrils, which then come together, and, depending on isoforms, either twist around each other or remain flat.

These results demonstrate that the fibrillization method of the invention is highly efficient for all Tau isoforms and is conducive to forming double-filament Tau fibrils, as exemplified by the TEM imaging of PK-treated 4R2N and PK-non-treated intact 3R2N fibrils of the invention (Figure 10c, arrows). The digestion with PK cleaves off “brush” residues on 4R2N fibrils and allows the two intertwined filaments to unwind. This thus far has not been demonstrated for conventional FFH- fibrillized 4R2N isoform using cryo-EM, whereas 3R2N doublets detected previously (Zhang et al., 2019, eLife 8, doi: 10. 7554/eLife.43584) appeared asymmetrical, in contrast to doublet fibrils found in this study, which were composed of the homomorphic singlet fibrils which then assembled into doublets. Example 5: Fibrillization of Tau isoform mixtures

The isoform composition of patient-derived pathological Tau fibrils is non-homogeneous and contains all six Tau isoforms, with ratios differing between the Tauopathies. Therefore, to confirm that the method of the invention would be suitable for modeling the complexity of Tau aggregation in the brain, the aggregation of four mixtures of Tau isoforms was assessed in varying compositions under the same conditions described above.

The mixtures comprised equimolar amounts of all six Tau isoforms [(All isoforms; 4R2N, 4R1N, 4R0N, 3R2N, 3R1N, 3R0N), isoforms containing 2N and IN repeats (2N + IN; 4R2N, 4R1N, 3R2N, 3R1N), isoforms containing 2N repeats (2N; 4R2N, 3R2N) and isoforms containing IN repeats (IN; 4R1N, 3R1N) in three individual reactions, (Reaction 1, 2 or 3) (Figure Ila)]. All samples showed efficient fibrillization, where Tau was enriched in the pellet fractions (Figure 11b). Interestingly, the relative incorporation of 4R-containing isoforms into the fibril-containing pellet fractions was consistently higher than 3R-containing isoforms (Figure 11c, Figure 12). Notably, the All isoforms (Reaction 2) showed a complete incorporation of all six Tau isoforms into the fibrils.

Next, the binding of the fibrils generated was assessed in the different mixtures to ThS (Figure lid). All samples showed a ThS signal indicating a formation of the amyloid motif-containing fibrils. Notably, the All isoforms (Reaction 2) demonstrated the highest ThS signal, indicating potential morphological or ultrastructural differences from all other isoform mixtures’ fibrils. To further assess the morphology of the fibrils, EM imaging was performed (Figure lie). All isoforms (Reaction 1) contained flexible fibrils of 15.15 ± 3.03 nm in width, and 3 contained similarly bendy fibrils of 15.08 ± 3.00 nm in width. The All isoforms (Reaction 2), however, showed a homogeneous population of the thin, rigid fibrils 11.02 ± 1.83 nm in width, significantly narrower than the All isoforms (Reaction 1) and the All isoforms (Reaction 3) (Figure Ilf, Table 2 below). 2N + IN (Reactions 1, 2 and 3) contained thick, flexible fibrils; similar to the 2N (Reactions 1, 2 and 3). IN (Reactions 1, 2 and 3), however, formed flat, right-hand twisting ribbon-like fibrils. The fibril width showed a strong negative correlation with the ThS signal, signifying those thinner, likely singlet fibrils have higher surface availability for the ThS dye binding (Figure 11g) and thus a high ThS binding/fluorescence.

Table 2 Together, these data show that the method of the invention is suitable for investigating the aggregation of Tau isoform mixtures and allows for the efficient production of Tau fibril preparations composed of mixtures of Tau isoforms, providing further opportunities to investigate variable ratios of non-modified and modified Tau isoforms that more closely represent isoform profiles in the human pathologies. Furthermore, this method allows minimizing the impact of differential FFH binding to isoforms, thereby influencing the dynamics of the isoform incorporation and co-aggregation in cases when heparin-binding affinity is higher for one or several Tau isoforms than the rest. In addition, the highly controlled in vitro aggregation method in the absence of the incorporated co-factors allows an in-depth characterization of the aggregation and morphological diversity of Tau fibrils produced.

Example 6: Fibrillization in the presence of co-factor molecules with a method according to the invention

Previous polyanion-based Tau aggregation methods did not allow for investigation into the effects of co-factors or other Tau ligands at different stages of Tau oligomerization and fibril formation because of competition with the polyanions in solution or their tight binding to Tau aggregates. The method of the invention addresses this limitation because the heparin is immobilized, and other co-factors or ligands could be added at different time points during the aggregation of Tau proteins without interference. Polynucleotide molecules, such as RNA (Kampers et al., 1996, FEBS letters 399, 344-349) and mononucleotide molecules, such as ATP (Farid et al., 2014, Microscopy research and technique 77, 133-137, doi: 10.1002/jemt.22319) contain a negative charge and have been used to induce aggregation of Tau in vitro. RNA has been detected associated with Tau- positive aggregates in human pathologies (Ginsberg et al., 1998, Acta Neuropathol., 96, 487; Ginsberg et al., 1997, Ann. Neurol, 41, 200). Furthermore, Tau-RNA interactions are important in liquid-liquid phase separation and formation of condensates (Boyko et al., 2022, Trends in Cell Biology, https://doi.Org/10.1016/j.tcb.2022.01.011). As shown previously, heparin could compete with the binding of RNA, displacing it from the growing Tau fibril (Fichou et al. 2018, Proceedings of the National Academy of Sciences of the United States of America 115, 13234-13239, doi: 10.1073/pnas.1810058115), likely due to partially-overlapping binding sites on Tau. Therefore, the interactions of these molecules were investigated, RNA and ATP, with Tau upon fibrillization, but in the absence of the free-floating polysaccharide heparin.

The aggregation of 4R2N Tau in the presence of other co-factors, polyuridylic acid (polyU, singlestranded RNA), or ATP by the method of the invention was compared to a standard method in the presence of FFH (1:4 ratio). The whole samples were imaged using EM midway through the reaction at 24 h and at the endpoint at 48 h (Figure 13a and c). Endpoint samples were fractionated to yield soluble supernatant (S) fractions containing monomeric Tau that was not incorporated into the fibrils, and pellet (P) fractions, containing fibrillized Tau. The samples were visualized using SDS-PAGE (Figure 13b and d).

In the presence of RNA, the EM imaging demonstrated the fibrillization of Tau by the ClearTau method, with numerous fibrils present at 24 h of the reaction (Figure 13a). FFH samples failed to aggregate at 24 h and showed indiscriminate amorphous aggregates and short fibrils at the later time point at 48 h (Figure 15a, arrows). In addition, the sedimentation assay showed efficient incorporation of Tau monomers into the pellet fraction (Figure 13b, ClearTau, grey arrows) in the presence of RNA, with a small amount of monomer present in the supernatant (Figure 13b, ClearTau, black arrows) for ClearTau samples. In contrast, the FFH samples showed low levels of incorporation of Tau into the pellet fraction (Figure 13b, FFH, grey arrows), with most protein detected in the supernatant (Figure 13b, FFH, black arrows). These observations illustrate the unique advantage of using the claimed method to investigate the interaction between Tau and other ligands or aggregation co-factors and highlight the limitations of the standard FFH method.

In the presence of ATP, both ClearTau and FFH samples showed the presence of the fibrils both at 24 h and 48 h of aggregation reactions (Figure 13c). The ClearTau fibrils appear long, curly, flexible and fuzzy. FFH samples show long, more flat fibrils. Again, the sedimentation assay demonstrated high levels of Tau incorporation into the pellet in ClearTau samples (Figure 13d, ClearTau, grey arrows) with a small amount of Tau remaining in the supernatant (Figure 13d, ClearTau, black arrows). The FFH samples showed high inter-sample variability, with the reaction FFH 1 demonstrating good incorporation of Tau into the pellet, whereas samples FFH 2 and FFH 3 showing lower incorporation of the protein into the fibril-containing fractions (Figure 13d, FFH, grey arrows). FFH 3 demonstrated a lower amount of protein in the pellet fraction (Figure 13d, FFH, black arrows).

These results demonstrate the higher reproducibility and efficient fibrillization of Tau in the presence of co-factors by the method of the invention. EM imaging showed extremely long fibrils (upwards of 5 pm) bundled into flexible cable-like structures and fuzzy appearance. Interestingly, biochemically the ClearTau samples appear to be more stable upon both RNA and ATP addition and show a higher amount of fibrillar Tau compared to co-factor-free reactions (Figure 13b and d, ClearTau). FFH samples demonstrated poor fibrillization in the presence of RNA and highly variable fibrillization in the presence of ATP. These results suggest potential interference of FFH with the co-factor or ligand molecules and their interactions impacting the Tau fibrillization propensity, as opposed to the method of the invention.

Example 7: Tau 4R2N P301L fibrils of the invention are twisted and bind RNA

The method of the invention was applied to aggregation of pathologically-relevant 4R2N Tau containing mutation P301L. P301L is associated with familial fronto-temporal dementia linked to chromosome 17 (Huton et al., 1998, Nature 393, 702), and P301L Tau readily forms fibrils in vitro in the presence of heparin, and in the method of the invention (Figure 6). Therefore, its fibrillization and biochemical, RNA-binding and ultrastructural properties were investigated with both ClearTau and FFH aggregation protocols.

From the EM imaging, the ClearTau P301L fibrils indistinctly showed periodical twisting, whereas the FFH P301L samples showed flatter fibrils (Figure 14b & c, -PK). It was hypothesized that the Tau N- and C-termini that remain disordered may obstruct the visualization of the ultrastructure of the ordered amyloid core of the fibrils. Thus, the limited proteolysis approach, that removes the protein parts distal to the fibrillar core (Figure 14a) was used using 10 pM proteinase K digestion of the fibrils for 1 min to cleave off the disordered “brush” residues. After digestion, the ClearTau fibrils demonstrated the sharp outline of the periodically-twisted fibrillar core (Figure 14b & c, +PK), whereas the FFH fibrils were straighter. Further, the quantifications of the fibril widths pre- and post-proteolysis showed ClearTau fibrils consistently wider than FFH fibrils, suggesting differing fibrillar core morphologies (Figure 14d). Biochemical fractionation by ultracentrifugation of the fibrils into soluble and pellet (insoluble) fractions and assessment using PAAG blot revealed a consistently higher amount of monomer incorporation into the fibrils for the ClearTau samples than FFH samples (Figure 14e, Figure 15a) since near-complete incorporation of Tau into pellet is observed for ClearTau and residual protein in the soluble fraction in FFH.

Finally, the RNA-binding properties of the fibrils prepared by both methods was assessed by incubating fibrils with RNA and assessing the presence of RNA in the soluble and fibril-bound fractions. The results showed significantly higher binding of the polyU RNA to the ClearTau fibrils in both buffer conditions, with the higher ionic buffer strength resulting in the substantial increase in RNA binding to the fibrils (Figure 14g). polyA RNA binding was not significantly different between the ClearTau and FFH samples, with low binding levels in the aggregation buffer and high binding in the presence of 100 mM MgCh. The yeast transfer RNA failed to bind either type of fibrils in aggregation buffer or in the presence of 100 mM MgQz (Figure 17b).

These results further strengthen the versatility of the method of the invention for use with different Tau proteoforms and in combination with downstream investigational methods. Protein expression and purification

All proteins were expressed and purified as described previously Ait-Bouziad et al., 2020, The Journal of biological chemistry 295, 7905-7922, doi:10.1074/jbc.RAl 19.012517).

Human 4R2N: Briefly, for the 4R2N isoform (P10636), the competent E. coli cells BL21 were transformed with plasmid pT7-7 SUMO-Tau human full-length 4R2N and incubated for 30 min on ice, heat-shocked for 45 sec at 42°C water bath, incubated on ice for 2 min. 300 pl of SOC outgrowth media (Thermo Fisher Scientific) were added, and the tube was incubated for 30 min @ 37°C with shaking. Cells were plated on Luria-Bertani (LB) solid agar broth with kanamycin plates and incubated overnight in a 37°C incubator for single colony growth. 20 ml of LB (Thermo Fisher Scientific) with 50 mg/L kanamycin were inoculated with a single bacterial colony and grown in the shaking incubator at 18°C overnight. To 8 L of autoclaved and filtered LB antibiotic kanamycin was added to the final concentration of 50 mg/L, media was split into 4 x 2 L flasks, inoculated with an overnight starter culture of SUMO-Tau 4R2N transformed bacteria, and grown at 37°C at 180 rpm in the shaking incubator to the confluence of 0.6-0.9 density. The culture was induced for the protein production by adding isopropyl P- d-1 -thiogalactopyranoside (IPTG, Thermo Fisher Scientific) to the final concentration of 0.4 mM and grown at 18°C overnight in the shaking incubator. The culture was decanted into 1 L centrifuge tubes and centrifuged at 3’000 rpm for 30 min. 150 ml of lysis buffer (50 mM Tris pH7.5, 30 mM imidazole, 500 mM NaCl) with 1 mM phenylmethylsulfonyl fluoride (PMSF) and protease inhibitor cocktail (SigmaAldrich) were added, bacterial pellet was solubilized fully, then sonicated on ice using probe sonicator using the following protocol: 70% amplitude, 30 s on, 30 s off for 5 minutes. Sonicated lysate was centrifuged for 30 min at 10 °C at 13’000 rpm. The supernatant was filtered and loaded on a HisTrap HP 5ml column (GE Healthcare). The purification was then performed using a linear elution gradient of His Trap Buffer A (50 mM Tris pH7.5, 30 mM imidazole, 500 mM NaCl) 100% to His Trap Buffer B (50 mM Tris pH7.5, 500 mM imidazole, 500 mM NaCl) 100%. The fractions were analyzed by SDS-PAGE, and pooled accordingly and cleaved by ULP1 enzyme (Thermo Fisher Scientific) for 1 hour at room temperature (RT). The pooled fractions were then loaded on a reverse-phase HPLC C4 column (PROTO 300 C4 10 pm, Higgins Analytical; buffer A: 0.1% TFA in water, buffer B: 0.1% TFA in acetonitrile), and the protein was eluted using a gradient from 20 to 40% buffer B over 90 min (15 ml/min). Fractions were analyzed by mass spectrometry and ultra-performance liquid chromatography for purity and pooled accordingly. Protein samples were flash-frozen in liquid nitrogen and placed into a vacuum lyophilizer for 48 h to produce lyophilized protein powder. All other Tau isoforms and variants were produced following a similar protocol. Fibril seed characterization

The fibrils obtained by the process described above were diluted to 10 pM in dITO and sonicated at 70 % amplitude for 50 s with 1 sec ON 1 sec OFF cycle in-tube using UP200St with VialTweeter (Hielscher, USA). These sonicated forms of Tau protein were designated “seeds”, and then were characterized by electron microscopy.

ThS fluorescence measurement

The fibrils obtained by the process described above were diluted to 2.5 pM in dFFO and sonicated at 70 % amplitude for 50 s with Is ON 1 sec OFF cycle in-tube using UP200St with VialTweeter (Hielscher, USA). 2.5 pM full-length Tau 4R2N monomer was used as a control. To the 100 pl reaction, 100 pl of ThS (10 pM) were added, yielding final protein concentrations of 1.25 pM. Single-timepoint ThS fluorescence was measured using 96 well clear bottom plate (Corning) set up in FLUOstar Omega microplate reader (BMG LABTECH, Germany) with excitation at 445 nm and emission at 485 nm was recorded. Four independent measurements were conducted in triplicates using ClearTau fibrils from two independent aggregation preparations. Raw fluorescence values were standardized to the blank reaction samples containing ThS. The plot represents the average of four experiments in triplicates, and bars show the standard deviation. Analogous measurement procedures were followed for other Tau isoforms and mixtures thereof, the details for each indicated in the legends.

ThS Tau seeding aggregation assay

Tau seeds were prepared from Tau fibrils produced using the aggregation protocol of the method according to the invention as described above. Reactions were set up in the clear bottom 96 well plates (Corning) as follows: Tau 4R2N monomer was diluted in the phosphate aggregation buffer (10 mM Phosphate, 50 mM NaF, 0.5 mM fresh DTT) to 10 pM. Tau seeds or free heparin sodium salt (Applichem GmbH) were added to the solution batches containing the Tau 4R2N monomer solution at a final concentration of 2.5 pM. The monomers without seed were incubated with the buffer as a control. A fresh ThS solution was added at 10 pM to the solutions and the plate was sealed with clear film. Reactions were conducted in FLUOstar Omega microplate reader (BMG LABTECH, Germany). The reading was taken from time 0 (corresponding to the maximum 20 min after the seed addition) every 10 min (1 cycle) for 19 h with shaking at 600 rpm for 10 min followed by the idle 10 min at 37°C. 4 independent replication experiments were performed in technical triplicates for each of the conditions listed above (1: Tau 4R2N monomers only; 2: ClearTau seeds with Tau 4R2N monomers; 3: FFH seeds with Tau 4R2N monomer). The Tau seeding values were standardized to the respective values derived from reaction containing only ClearTau or FFH seeds, but no monomer, heparin seeding values were standardized to the reaction containing heparin only and no monomer, monomer only reaction values were standardized to the reaction containing buffer and ThS only. The plot represents average values for 4 experiments, and bars represent the standard deviation.

Negative stain transmission electron microscopy (EM)

2 pl of samples in solution were deposited on a glow-discharged Formvar/carbon-coated 200-mesh copper grids (Electron Microscopy Sciences) for electron microscopy, incubated for 5 min, washed three times in dHjO, and stained using 2% uranyl formate solution. Images were acquired using a Tecnai Spirit BioTWIN transmission electron microscope operated at 80 kV and equipped with an LaB6 filament and a 4K x 4K FEI Eagle CCD camera.

Width quantification of Tau fibrils according to the invention

The Tau fibrils of the invention and FFH fibrils widths were quantified using EM images from at least 3 independent fibril preparations and independent EM grid preparations. Fibril widths were measured using ImageJ (Abramoff et al., 2004, Biophotonics international 11, 36-42) Measurement Tool (ImageJ, RRID:SCR_003070).

Monomer - Fibril fractionation

The ClearTau or FFH aggregation reactions were ultracentrifuged using Beckman-Coulter ultracentrifuge at 100’000 g for 1 h at 4°C. The supernatant was removed, the remaining pellet was washed in dH2O twice and then resuspended in dH2O.

SDS-PAGE protein assay

Twenty-five pg of total protein per well was loaded on fixed polyacrylamide concentration of 15 % SDS-PAGE gels (Invitrogen, Thermo Fisher Scientific) and run in MES buffer (Invitrogen, Thermo Fisher Scientific). The total protein content was visualized using Coomassie protein stain.

Biosensor (BS) cellular flow cytometry assay

Cell line Tau RD P301S Tau FRET Biosensor (Furman & Diamond, in Tau Protein 349-359, Springer, 2017) (CRL-3275™) was acquired from ATCC® and maintained in DMEM medium with 0.5 % L-Glutamine, 0.5 % penicillin-streptomycin antibiotic cocktail and 10 % foetal bovine serum supplementation (Gibco, Thermo Fisher Scientific). BS cells were plated in poly-L-lysine treated 6 well plates at a density of 100’000 cell/well. Cells were allowed to grow and divide in an incubator at 37°C. BS cultures were allowed to reach a confluency of 50-60%. The ClearTau sonicated fibril seeds were incubated with Lipofectamine2000 at 1:2 ratio weight to volume in OptiMEM (Gibco, Thermo Fisher Scientific). The cultures were transduced with ClearTau sonicated fibril seeds at amounts of 0.892 pg, 0.446 pg and 0.223 pg of fibrils per 200’000 cells. Cells were exposed to the fibrils for 4 h, cultures were washed twice in phosphate-buffered saline (PBS, Gibco, Thermo Fisher Scientific) to remove all residual seeds, and further incubated for 96 h in standard medium to allow to have two cell division cycles. Cultures were washed once in PBS, and dissociated using 200 pl Trypsin-EDTA 0.05% for 5 min at 37°C. 150 pl of DMEM medium was added to each well to neutralize Trypsin action, cells were gently dissociated into single cells by pipetting, they were transferred into Eppendorf tubes and centrifuged at 1000g at room temperature for 5 min. The supernatant was removed, and cells were re-suspended in the 900 pl of 2% paraformaldehyde (Thermo Fisher Scientific) and incubated for 10 min, then pelleted by centrifugation at 1000 g 4°C for 5 min. The supernatant was removed, and the pellet was resuspended in HBSS.

FRET Flow cytometry protocol was adjusted from (Furman & Diamond, in Tau Protein 349- 359, Springer, 2017). FRET detection was performed using BD LSR Fortessa with excitationemission laser filters: 405 - 465/30, 488 - 530/30, with FRET signal detectable at 405 - 530/30 couple. Parental HEK293T cells were used to define the cell population on the SC-A vs FCS-A bivariate plot. Doublet events were excluded on FSC-H vs. FSC-A bivariate plot. Voltages were adjusted to exclude any signal on CFP, YFP, or FRET filters. Double-positive CFP-YFP BS cell population was defined by the negative Lipofectamine-only BS cell line control sample. Compensation was adjusted to remove any bleed-through of CFP and YFP signal to the FRET channel (Figure 16).

For data analysis, cell populations were gated to exclude the debris events and doublets. Negative control (BS-Lipofectamine) was used to define double-positive cell population, spill-over to the FRET channel was excluded on CFP-FRET and YFP-FRET bivariate plots. For each sample, percent of FRET-positive events and Median of fluorescence intensity were recorded, and the product was plotted to represent Integrated FRET Density (IFD). Three independent experiments were performed for each condition with a minimum of 100’000 events per sample recorded. The plot represents average measurements; bars represent standard deviation.

Confocal imaging

For confocal imaging, Tau RD P301S FRET Biosensor cells were plated at a low density (>5’000/well) in the 24 well plates on the poly-l-lysine-coated (3438-100-01, R&D Systems) glass coverslips and allowed to attach overnight. Cells were incubated in the HBSS (Gibco) overnight, and then the medium was changed to normal, cells were allowed to double. Media was aspirated, cells were washed in IX PBS, 10 pM ClearTau preformed seeds were added in OptiMEM (Gibco) for 4 h, then the media was aspirated and replaced by the standard media. Cells were allowed to have two cell divisions, after which they were washed in IX PBS, and were fixed using 4% formaldehyde solution for 20 min. Cells were washed twice in IX PBS, and coverslips were mounted onto microscopy slides using Molwiol mounting medium (Sigma-Aldrich). Confocal images were acquired using Zeiss LSM 700 microscope.

Circular dichroism (CD) spectroscopy

CD spectra were recorded on a Jasco J-815 CD spectrometer operated at 20°C. To minimize buffer absorption, the samples were diluted with 1:50 deionized H2O. CD spectra were acquired from 190 nm to 260 nm at a scan rate of 50 nm/min and in increments of 0.2 nm. For each sample, three to four spectra were averaged and smoothed using binomial approximation.

Cryoelectron microscopy

The fibrils obtained by the method of the invention from Tau 3R2N and 4R2N proteins were screened with negative staining (NS) TEM for fibrils concentration and morphology. Aliquots of optimized fibril samples were applied onto gold Ultrafoil 1.2/1.3 grids, and plunge frozen in liquid ethane. Frozen cryo-EM grids were imaged on a ThermoScientific 200kV Glacios on K3 electron counting direct detection camera (Gatan Inc.) in counted (non-CDS) mode (50 fractions) using the SerialEM (Mastronarde et al., 2005, J Struct Biol, 152, 36-51) at a physical pixel size 1.113 A for 3R2N fibrils and 0.878 A for 4R2N fibrils, and a total dose of 50 electrons per square angstrom (e- /A2) for each exposure.

After inspection, best 6021 (3R2N) and 3331 (4R2N) aligned, CTF-estimated and dose-weighted movies were selected from FOCUS (Biyani et al, 2017, J Struct Biol, 198, 124-133) for further processing. Several hundreds of representative non-overlapping filaments were manually selected using the e2helixboxer.py from EMAN2 (Tang et al, 2007, J Struct Biol, 157, 38-46). Dose- weighted averages were denoised with JANNI, and subjected for the semi-automated filaments tracing with crYOLOvl.7.5 (Wagner et al., 2019, Commun Biol 2, 1-13). Filaments start-end coordinates in STAR format were imported into RELIONv3.1 (Zivanov et al., 2018, RELION-3. eLife 7, e42166), and 1,842,783 3R2N and 570,876 4R2N segments were extracted with a box size of 300 pixels and subjected for reference-free 2D classification in cryoSPARCv3.2 (Punjani et al., 2017, Nature methods, 14, 290-296). Several rounds of 2D classifications allowed to select only particles with clear 4.77 A beta-strand separation along the fibril axis, measured from Fourier amplitudes of the 2D class average. To separate singlet and doublet fibrils, helical segments were re-extracted with a larger box size of 900 pixels, re-scaled to 300 pixels and subjected for reference- free 2D classification in cryoSPARC. 2D class averages corresponding to singlets and doublets were separated and further classified.

Co-factor aggregation

To aggregate 4R2N Tau in the presence of co-factor polyuridylic acid single-stranded RNA (P9528, Sigma-Aldrich), or adenosine triphosphate (987-65-5, SERVA/Cayman chemical) by the method of the invention or in the presence of free-floating heparin (FFH), Tau 4R2N was added at 10 pg/ml to 1 mM of co-factor molecules. FFH was added at 1:4 ratio. The reactions were setup in triplicates and incubated at 37°C with orbital shaking at 1000 rpm for 48 h. 100 pl of endpoint reactions (W) were ultracentrifuged using Beckman-Coulter ultracentrifuge at 100’000 g for 30 min at 4°C. The supernatant (SN) was removed, remaining pellet (P) was washed in PBS twice, then resuspended in 100 pl PBS to yield a fraction containing fibrils. 10 pl of the fractions were mixed with 10 pl 2X Laemmli buffer, and 2 pl per well were loaded on the SDS-PAGE as described above.

Proteinase K digestion of fibrils

10 pM of Tau 4R2N or 4R2N P301L was fibrillized by a method of the invention or in the presence of FFH for 24 h. 300 pl of samples were ultracentrifuged, the supernatant removed and the pellet resuspended in 300 ul of PBS. 50 pl of samples were digested by proteinase K (39450-01-6, Invitrogen) at 10 pg/ml for 1 min. The reactions were quenched by the PMSF at 0.3 mM. The samples were imaged using electron microscopy as described above.

RNA-binding assay

Tau 4R2N P301L was fibrillized by a method of the invention or the presence of FFH for 24 h. 300 pl of samples were ultracentrifuged, the supernatant removed, the pellet resuspended in 300 pl of PBS. A total of 10 pg/mL of polyU RNA (P9528, Sigma- Aldrich), polyA RNA (P9403, Sigma- Aldrich) or yeast tRNA (R1753, Sigma-Aldrich) were added to 100 pL of 5 pM fibrils in the aggregation assay buffer (Phosphate buffer, pH 7.4). The fibril/RNA-mixture was incubated at 37 °C with 550 rpm shaking for 1.5 h. After the incubation, the 80 pL of fibril/RNA-mixtures was centrifuged at 160’000 g at 37 °C for 30 min in an ultracentrifuge. The supernatant was removed, and the pellet was resuspended in 80 pL of aggregation assay buffer with 2% SDS. The concentration of RNA in the supernatant and the pellet (resuspended) was calculated using a Nanodrop One spectrophotometer (Thermo Fischer Scientific) (Chakraborty et al., 2021, Nat Commun 12, 4231, doi: 10.1038/s41467-021-24362-8). Spectra were baseline corrected using the buffer as reference. The concentrations of RNA were calculated from a sample of 10 pg/mL RNA in 100 pL of aggregation buffer. Measurements were performed with three independently prepared samples in each case. Statistical analyses and data visualizations

All statistical analyses and data visualizations were performed using GraphPad Prism 9 software (San Diego, USA) and Microsoft Office Suite tools (USA). Atomic visualizations were performed using PyMOL software (Schrodinger, LLC. The PyMOL Molecular Graphics System, Version 1.8, 2015).

Brain and HEK293T cell lysate extract preparation

Lysate buffer was prepared in PBS with 1% PMSF, 1% phosphatase inhibitor cocktails 2 and 3, 1% PI. The buffer was filtered with a 0.1 um filter syringe. 50 pL of buffer were added to a mouse brain cortex hemisphere, or 100’000 HEK293T cells. The samples were sonicated on ice for 4 min (1 s ON, 1 s OFF, 60% amplitude), then ultracentrifuged at 100’000 g and 4°C for 30 min. Supernatants were removed and retained as lysates. 100 pM of Tau protein was added to 250 pl of Phosphate aggregation buffer. 10 pl of 0.6 mg/ml of mouse cortex soluble lysate or 10 pl of 1.9 mg/ml of HEK293 cell soluble lysate was added to the reaction. Reactions were incubated for 3 days with 1000 rpm shaking.

Nitration reactions

The Tau aggregates (fibrils) were incubated with 10 equivalent (eq.) of tetranitromethane (TNM, 10% in ethanol) per tyrosine residue to protein (2 mg/mL) keeping constant 5 tyrosine residues for all proteins and tyrosine mutants in sterile IX PBS buffer, pH 7.4 (Gibco) for 2h at room temperature. After the reaction TNM was removed by ultracentrifugation at 100’000 g for 30 min at 4°C, where the residual TNM in the supernatant was removed, and fibrils were washed twice in sterile PBS.

Example 8: Screening of Tau aggregation modulators

A method according to the invention can be used to screen and identify modulators of Tau oligomerization and fibril formation. (Figure 17a). An important timepoint in Tau aggregation is the initiation of its polymerization, and targeting this step is crucial for finding agents that can either stabilize Tau in its monomeric form or interfere with the polymerization and fibril growth process. An assay for screening of Tau aggregation modulators such as inhibitors, enhancers or structure modifiers is implemented as follows:

- Mixing together Tau monomers and at least one candidate modulator in a buffer solution and incubating the mixture for at least 24h;

Adding the obtained mixture into at least one reaction vessel with its reaction surface precoated with a co-factor covalently immobilized on-to this surface;

Incubating the obtained reaction mixture for at least 24h. Then, the incubated mixture is separated into two fractions i) one containing non-aggregated Tau monomers (“Soluble”) and ii) one containing aggregated Tau species, such as fibrils or oligomers determined by electron microscopy (EM) imaging and biochemical profile by SDS-PAGE (“Pellet”).

The intensity ratios of Soluble to Pellet fractions indicate whether the candidate modulator stabilized the Tau monomer and prevented Tau aggregation (Figure 17a, I: Tau aggregation inhibitor, “compound A“), whether the candidate modulator increased the aggregation of Tau (Figure 17a, II: Tau aggregation enhancer, “compound B“), or whether the candidate modulator induced formation of oligomeric Tau species (Figure 17a, III: Tau oligomerization enhancer, “compound C“).

Morphological examination by EM imaging reveals the fibril-remodelling properties of the screened candidate modulator (Figure 17a, IV: Tau aggregate morphology modifier, “compound D, E, F“).

The reaction yields high amounts of pure compound- or compound-induced Tau proteoforms, such as fibrils and oligomers, that can be further studied for their biochemical properties in vitro, cytotoxicity and seeding in cell cultures, and neurotoxicity, seeding and spreading in the brain injection animal models.

As validation test, a panel of known and unknown modulators and/or inhibitors of Tau aggregation were used as candidate agents and then classified according to their effect profile as described in Figure 17a:

I. Tau aggregation inhibitor

II. Tau aggregation enhancer

III. Tau oligomerization enhancer

IV. Tau aggregate morphology modifier.

For implementing the above assay, monomeric human full-length 4R2N Tau was obtained as described above, and was diluted to 10 pM in phosphate buffer (10 mM Phosphate, 50 mM NaF, 0.5 mM fresh DTT) to obtain a suspension of the monomers in aqueous solution in an Eppendorf tube.

10 pM of candidate agents ATPZ (Cpdl6 - Calbiochem, 5-amino-3-(4-chlorophenyl)-N- cy cl opropyl-4-oxo-3,4-dihydrothieno[3,4-d]pyridazine-l -carboxamide; 580222, Sigma), pyrocatechol violet (P7884, Sigma), BSc3094 (2-[4-(4-Nitrophenyl)-2-thiazolyl]hydrazide-lH- benzimidazole-6-carboxylic acid monohydrobromide; B7937, Sigma), LMTX (Hydromethylthionine mesylate S7762, Selleckchem), myricetin (3,3,4,5,5,7-Hexahydroxy- flavon, Cannabiscetin, Myricetol; M6760, Sigma), DMSO (Methyl Sulfoxide, Methyl sulfinylmethane; D8418, Sigma), dopamine (2-(3,4-Dihydroxyphenyl)ethylamine hydrochloride, 3,4-Dihydroxyphenethylamine hydrochloride, 3-Hydroxytyramine hydrochloride, 4-(2-Aminoethyl)-l,2-benzenediol hydrochloride; H8502, Sigma), L-DOPA (3-(3,4- Dihydroxyphenyl)-L-alanine, L-3-Hydroxytyrosine, Levodopa; D9628, Sigma), 4-HNE (4- hydroxy-2E-nonenal; 32100, Cayman), EGCG (Epigallocatechin gallate, (-)-cis-2-(3,4,5- Trihydroxyphenyl)-3,4-dihydro-l(2H)-benzopyran-3,5,7-triol 3-gallate, (-)-cis-3,3 ' ,4 ' ,5,5 ' ,7-Hexahydroxy-flavane-3-gallate; E4143, Sigma) were added to the monomer suspension, resulting in a 1:1 molar ratio and incubated for about 48h.

Then, 200 pl of the mixtures were then transferred Eppendorf tubes with a co-factor (e.g. heparin) covalently immobilized on-to their surface, and the tubes were placed on an orbital shaker at 1000 rpm at 37°C for 48h aggregation.

The reaction mixtures were then ultracentrifuged using Beckman-Coulter ultracentrifuge at 100’000 g for 1 h at 4°C, the supernatant was removed and saved as a Soluble fraction, the pellet was washed twice with dFLO. The pellet containing fibrils was resuspended in dFLO to 200 pl and labelled Pellet fraction. For EM, 5 pl of the Pellet were loaded as described below. For SDS-PAGE, to 10 pl of samples 10 pl of Laemmli buffer were added, and 5 or 10 pl of each sample was loaded on 15/5% PAA gel. Total protein was visualized by Coomassie stain. The quantification of the soluble and pellet fraction ratios was performed using ImageJ (Abramoff et al., 2004, supra) Gel Tool (ImageJ, RRID:SCR_003070). The observed effects on the fibril morphology by some of the candidate agents as observed by EM are presented under Table 3 below.

Table 3

As can be seen on Figure 17, b, c an aldehyde 4-HNE and catechin EGCG compounds were found most effective at preventing Tau fibril formation and induced the formation of the oligomeric species as detected by EM (Figure 17, e). 4-HNE and EGCG correspond to compounds with profiles I. Tau aggregation inhibitor, HI. Tau oligomerization enhancer and IV. Tau aggregate morphology modifier, with a distinct ability to promote Tau oligomerization.

As a validation, LMTX which is a reduced form of methylene blue, well-studied as a Tau aggregation inhibitor (Wischik et al. 1996, PC, Proc Natl Acad Sci USA, 93: 11213-11218) did not fully prevent fibril formation in the assay, however most of the Tau protein indeed remained in the Soluble fraction (Figure 17 b, c), indicating that LTMX acts as stabilizing the soluble forms of the protein. LMTX had a pronounced impact on the observed morphology and stability of the fibrils that were still formed. Fibrils were fragmented and had an aberrant morphology, but only few oligomers were also present. Similar to 4-HNE and EGCG, LMTX belongs to categories I. Tau aggregation inhibitor and IV. Tau aggregate morphology modifier, but without pronounced oligomer-inducing properties.

ATPZ, pyrocatechol violet, BSc3094, and myricetin did not prevent Tau fibril formation. In fact, all but BSc3094 increased the aggregation of Tau to varying levels, with myricetin being most potent. All compounds had pronounced effects on the fibril morphologies, with fibril widths significantly smaller than controls (Figure 17, d, Table 3). Morphologically, ATPZ and BSc3094 induced formation of multiple fibril conformers and high amount of fragmentation. On the other hand, pyrocatechol violet and myricetin induced formation of uniform populations of thin fibrils, with wavy conformers in pyrocatechol violet, and sharply-defined thin fibrils in myricetin reactions (Figure 17c, d). These compounds most fit the profiles of varying-level II. Tau aggregation enhancer and IV. Tau aggregate morphology modifiers.

DMSO also did not prevent Tau fibril formation, but had slightly increased it (Figure 17b, c). Morphologically, DMSO-induced fibrils were not different from controls, classifying it as a low- level II. Tau aggregation enhancer.

L-DOPA is a precursor to dopamine, and due to its extra carboxy group, maintains zwitterionic profile of charge. Similar to compounds ATPZ, pyrocatechol violet, BSc3094, myricetin and DMSO, L-DOPA enhanced the shift of Tau into Pellet fraction, with only a minority of protein remaining in the soluble state (Figure 17b, c). The fibrils formed in L-DOPA presence were homogenous, however they were thinner and straighter than control fibrils (Figure 17d, e). L- DOPA also corresponds to II. Tau aggregation enhancer and IV. Tau aggregate morphology modifier profiles.

Dopamine is formed by the loss of carboxy group of L-DOPA, which confers a positive charge profile onto the molecule. Morphologically, the Tau fibrils formed in the presence of dopamine were in stark contrast to L-DOPA fibrils, showing wide cable-like bendy fibrils without clear definition of the fibril surface. The oligomeric species were also present, classifying dopamine as III. Tau oligomerization enhancer and IV. Tau aggregate morphology modifier.

Those results are in line with those obtained for known Tau aggregation inhibiting compound, such as LMTX. Importantly, it provides new insights on the action of compounds ATPZ, pyrocathechol violet, BSc3094, myricetin, L-DOPA, 4-HNE and DMSO on Tau aggregation at biochemical and structural levels.

In summary, the methods of the invention offer the possibility to design effective assays to discern specific effects of molecules of interest on the de novo formation of Tau fibrils, a crucial step of the Tau aggregation cascade of events. These effects include morphological changes of fibrils as compared to control no-intervention samples, such as whether multiple conformers are present within the mixture (homogenous vs. heterogenous), width and fibril shapes and fragmentation, as well as changes in biochemical profile from soluble to insoluble. Further, the methods and assays according to the invention allows to identify the propensity of an agent to inhibit or enhance fibrillization, and/or induce the oligomer formation. Importantly, the methods and assays according to the invention allow the generation and purification of compound-induced Tau proteoforms, such as fibrils and oligomers that can be further studied for their biochemical properties in vitro, cytotoxicity and seeding in cell cultures, and neurotoxicity, seeding and spreading in the brain injection animal models as a useful research tool.

Example 9: Tailored preparation of pathology-relevant Tau

The method of the invention can be applied to prepare various range of pathology-relevant Tau protein aggregates in the instances where the biochemical aggregation conditions would lead to formation of desired aggregate morphologies, and these aggregates can then be fully characterized and verified in downstream tests to accurately represent the pathological structures found in brain. Therefore, a fine-tuning of the aggregation conditions (relevant co-factors and molecules, shaking speed, volume, temperature, pH etc.) are possible to lead to recapitulation of specific pathological aggregate structures. The following examples are provided below to support the versatility of the possible applications of the invention. a. Production of more tightly-twisting, thick and very long (5+ m) Tau fibrils (Figure 18a). The present invention allows for the modulation of the fibril twist conformation. b. Production of Tau fibrils in the presence of soluble co-factor extracts from cortical mouse brain and HEK293T cell line. The co-factors and multiple molecules present in the brain may impact the morphologies and properties of the growing Tau fibrils once the nucleation is triggered. The method according to the invention can be used to assess the impact of the cell-derived cofactors on the Tau fibril morphologies and remodeling. 4R2N Tau monomers were aggregated using the method of the invention in presence of brain and cell line lysates. The brain extract-induced fibrils show twisting singlet fibrils with thick “brush” density. HEK293T-induced fibrils are predominantly flat, with twisting present at wider intervals along the fibril, and thin outer density (Figure 18b). c. Production of multiple tyrosine-to-phenylalanine mutant Tau aggregates for anti-nitrated or anti-phosphorylated tyrosine antibody development. Five tyrosine residues in Tau protein are of particular pathological relevance with regards to their post-translational modifications, such as nitration and phosphorylation. Three tyrosines are located in the C- terminus close to (Y197 and Y384) or within (Y310) the fibril amyloid core. Mutations or post-translational modifications of these residues may disrupt and lower the aggregation efficiency and kinetics, therefore it is important to investigate their precise functions. To demonstrate the applicability of the invention to these cases, Tau tyrosine mutants were aggregated using the method of the invention method. 4R2N and other full-length isoforms contain five tyrosines, whereas the tyrosine mutants include proteins containing a single tyrosine at Y18 or Y29; two tyrosines at both Y18 and Y29; three tyrosines at Y197/Y310/Y394 with phenylalanines at residues 18 and 29; and four tyrosines with either Y18 or Y29 mutated to phenylalanine. Efficient aggregation of the Tau tyrosine mutant proteins into fibrils was achieved (Figure 18c, naive), which are at high quantity and pure quality for the use for downstream applications, for example nitration of such aggregates (see Example 9d), and other post-aggregation modifications. This demonstrates that the method is translatable to a wide variety of Tau isoforms and mutants, and yield a substrate for the further usage in biochemical studies (see Example 9d). d. Production of nitrated wild-type and mutant ClearTau aggregates (Figure 18c). Nitration of pathological protein aggregates has been detected and implicated in the progression of neurodegenerative disorders, such as Alzheimer’s disease. Therefore, it is necessary to investigate its impact on the aggregated Tau protein species, specifically in the aggregate- disrupting/disaggregating and/or aggregate-modifying role. Tyrosine residues are the predominant sites of nitration reaction, therefore the aforementioned Tau tyrosine mutant fibrils (Example 9c) along with non-mutant Tau isoforms 3R2N, 4R0N, 4R1N and 4R2N were used as substrates for nitration. The superiority of using Tau fibrils of the invention over conventional heparin-induced fibrils is the lack of any residual carbohydrates that could react with the strong nitrating agent, such as tetranitromethane, impacting the nitration reaction of proteins and introducing artifacts. The results demonstrate the post- fibrillization remodeling of the nitrated Tau fibrils of the invention, such as extensive fragmentation into short stubby fibrils. These results demonstrate the utility of production of desired Tau aggregate morphologies, and the versatility of the method and co-factor free fibrils of the invention for addressing various scientific questions and hypotheses testing.

Example 10: Stability of the efficacy of co-factors immobilized on walls of a reaction vessel after multiple rounds of aggregation reactions

The aggregation vessels on the walls of which co-factors are covalently immobilized are reusable over multiple rounds of aggregation, which is economical and waste-reducing solution for plastics in the laboratory setting. However, it is important to validate and characterize the resultant Tau aggregates after each re-use cycle to ensure no loss of aggregation efficiency or changes in morphology of said aggregates. A heparin-coated reaction tube was provided (ThermoFisherScientific) where heparin was immobilized by amide coupling without linker as described in Linhardt et al., 2008, supra). To show that the method of the invention can be used to aggregate protein on the large scale with minimal amount of investment, high efficiency and morphological consistency, the same aggregation tubes were vigorously washed for up to five times between the aggregation reactions using PBS and detergent Hellmanex at 1% (Figure 19a), resulting in six rounds of pure co-factor free Tau fibril preparations of full-length Tau isoforms 3R2N and 4R2N with mutation of residue 18 tyrosine to phenylalanine (4R2N/Y I 8F ). The gel and EM demonstrate the consistent shift of soluble protein into insoluble fibrillar fractions (Figure 19b, arrows), and formation of fibrillar aggregates after wash 5 for both isoforms.

This shows that the method of the invention is easily translated into practical use in the large-scale and automated production of the Tau aggregates with consistent output.

Altogether, those data support that the method of the invention and the resulting Tau fibrils can be used in a variety of applications in drug development and research as illustrated in Figure 20 and stress its compatibility with emerging technologies such as organoids and other precision methods that could be very sensitive to impurities.