An aggregation-induced emission-active theranostic agent for selectively detecting and intervening pathological Tau protein

Abstract

The accumulation of Tau aggregates is commonly linked with various neurodegenerative diseases, such as Alzheimer's disease, Pick's disease, and corticobasal degeneration. Notwithstanding substantial investments in the development of clinical strategies for effective intervention, traditional design paradigms are predominantly confined to molecules featuring either a solitary function or single-dimensional mode of intervention, ignoring the necessity of personalized and precise medicine. Herein, we design and synthesize a dual-functional aggregation-induced emission-active agent to serve as both a fluorescent probe for the imaging of pathological Tau and a modulator for intervention. This amphiphilic theranostic agent, named TPE-P9, is prepared via a one-pot Michael reaction between hydrophobic maleimide-modified tetraphenylethylene (TPE-Mal) and a hydrophilic cysteine-modified Tau-targeting peptide (CKVQIINKK). Microscale thermophoresis measurement and in vitro fluorescence analysis demonstrate that TPE-P9 exhibits specific binding affinity (K_d = 4.46 µM) and high selectivity towards Tau fibrils, featuring a pronounced low background interference, which is superior to the classical amyloid protein probe thioflavin T (ThT). At the living cellular level, TPE-P9 is capable of readily imaging endogenic pathological Tau to distinguish normal neurons from the lesional neurons in situ, and the staining consequence is almost consistent with that of ThT. On the other hand, as a modulator, TPE-P9 can potently protect neurons from cytotoxic Tau-induced apoptosis both by inhibiting aberrant post-translational modification-induced Tau self-assembly and by blocking the produced pathological Tau propagation, enhancing cell viability by 35.4%. These findings offer valuable insights for the development of innovative image-guided therapeutic strategies for targeted tauopathies treatment.

---

Introduction

Tau protein, a microtubule-associated protein abundantly expressed in diverse neurons, presents in six isoforms within the adult human brain. These isoforms are categorized by the presence or absence of N-terminal inserts (N1, N2) and a C-terminal repeat domain (R2), with 2N4R Tau being the longest at 441 amino acid residues.^1,2 Native Tau interacts with microtubules to promote microtubule assembly and stability through its microtubule-binding domain (K18) that comprises three or four repeat domains. The various functions of Tau are mainly modulated by protein post-translational modifications (PTMs) including phosphorylation, ubiquitination, acetylation, methylation, and other processes.^3,4 In the pathological state, aberrant PTMs weaken Tau's microtubule-binding ability, leading to microtubule depolymerization, mitochondrial dysfunction, synaptic plasticity impairment, and others.⁵ These produced abnormal Tau proteins are released from pathological neurons and spread among cells in a ‘prion-like’ manner. The intracellular misfolded Tau accumulates and tends to aggregate into oligomers, protofilaments and filaments in neuronal disorders termed as tauopathies, such as Alzheimer's disease (AD), Pick's disease (PiD), progressive supranuclear palsy (PSP), and more.⁶

The accumulation of β-sheet-rich amyloid aggregates is generally regarded as the pathological hallmark of many neurodegenerative diseases, promoting extensive research into suppressors and degradative agents for these toxic species. Aducanumab, lecanemab, and donanemab have been approved by the U.S. Food and Drug Administration (FDA) as new therapies for AD by alleviating amyloid beta (Aβ) plaque burden.^7–9 A monoclonal antibody, mAb47, was reported to be capable of reducing levels of α-synuclein protofibrils in the spinal cord and improving motor function in transgenic mice.¹⁰ Unfortunately, as of now, none of the specific drugs targeting pathological Tau are approved for listing, and the studies on such therapeutic strategies remain insufficient. Emerging pieces of evidence have pointed out that Tau tangles might exert more prominent impacts on the progress of neurodegenerative diseases than Aβ plaques.^11–13 However, certain reported molecules against tauopathies have demonstrated limited efficacy in clinical implementation, primarily due to several constraints including uncertain physiological toxicity, poor blood–brain barrier (BBB) penetration, inadequate targeting specificity, monofunctional property, and molecular heterogeneity. Moreover, some studies utilized truncated protein fragments instead of full-length Tau to develop pathological models for laboratory research,^14,15 failing to consider the impact of extended motifs beyond these selected fragments on the protein's overall conformation and the accessibility of its active sites. Therefore, it is a pressing need to design and develop a highly efficacious and specific strategy targeting full-length pathological Tau protein for the early diagnosis and therapeutics of tauopathies.

From the perspective of disease diagnosis, fluorescence-based probes attract increasing attention owing to their merits in rapid response time, high sensitivity, and low cost. For example, Li et al. modified gold nanostars with multicolor fluorescence-labeled nucleic acids to achieve real-time and in situ fluorescence imaging of apoptosis pathway cascades.¹⁶ A class of DNA-templated gold nanoparticle-quantum dot assembly-based probes has been applied for facile identification of cancer cells.¹⁷ Additionally, a metal–organic framework (MOF) nanoprobe doped with zirconium and boric acid was proficient in simultaneous detection of protein phosphorylation and glycosylation levels in vivo.¹⁸ However, the inherent defects of traditional fluorescent molecules, such as the aggregation-caused quenching (ACQ) effect and inevitable noise from an always-on pattern, may severely restrict their signal response ranges and diminish the detection sensitivity.^19,20 To address these issues, Tang and co-workers proposed a series of novel aggregation-induced emission (AIE) molecules in 2001.²¹ In contrast to conventional luminescent materials, AIE luminogens (AIEgens) exhibit minimal emission when dispersed but demonstrate intense luminescence in aggregated or solid states, attributed to the restricted intramolecular motion. This characteristic facilitates AIEgens’ effectiveness as ideal tools for detecting amyloid proteins with intrinsic aggregation tendencies, as their emission activity is intrinsically aligned with the aggregation degree of target amyloid proteins.^22,23 However, common AIEgen core units, including tetraphenylethene (TPE), tetraphenylpyrazine (TPP),²⁴ and triphenylamine (TPA),²⁵ are nonspecific and hydrophobic, leading to susceptibility to interference and false-positive signals from self-aggregation in aqueous detection. Therefore, additional structural modifications are required to achieve their practical applications further.

The use of recognition motifs, such as antibodies, aptamers, and peptides, is a promising strategy for targeting amyloid proteins. Peptides, in particular, offer notable advantages over the other two biomolecules, including comparable or superior target affinity and selectivity, enhanced stability, minimal batch-to-batch variability, simplified quality control, and reduced manufacturing and shipping costs. Currently, peptide-based nanomaterials are extensively employed in tumor diagnostics,²⁶ protein activity assessment,²⁷ theranostics for infected cells,²⁸ and a range of other biomedical applications.

In this study, we designed a novel dual-functional probe, 1-(4-(1,2,2-triphenylvinyl)phenyl)-1H-pyrrole-2,5-dione-CKVQIINKK (TPE-P9), via peptide and AIEgen covalent conjugation to achieve simultaneous Tau protein detection and aggregation modulation. Integrating imaging and intervention within a single compound enables synchronized real-time accurate diagnosis and in situ effective treatment, optimizing the intervention window while concurrently monitoring drug distribution and evaluating therapeutic efficacy, thereby advancing the objective of personalized and precision medicine.²⁹

The conjugation of P9 with AIEgen synergistically combines the merits of sequence-based and fluorescence-based probes, improving selective targeting, therapeutic efficacy, aqueous solubility, biocompatibility, and biological imaging capability. As an amyloid probe, TPE-P9 exhibited turn-on fluorescence upon interaction with Tau aggregates and distinguished various Tau aggregation statuses through different fluorescence intensities (Fig. 1a and b). Compared to the widely used β-sheet fluorophore dye thioflavin T (ThT), this probe demonstrated significantly greater binding affinity and selectivity for fibrillar Tau. As an amyloid modulator, TPE-P9 regulated the aggregation process of pathological Tau through two distinct mechanisms, either by acting as a steric zipper inhibitor or by forming cross-seed (Fig. 1c and d). This modulation mitigated pathological Tau-induced neuronal apoptosis at various stages, positioning TPE-P9 as a promising candidate for tauopathies treatment. To the best of our knowledge, this is the first report of a dual-functional peptide-AIE conjugation system targeting Tau protein, enabling simultaneous monitoring, detection, and intervention in Tau fibrillogenesis. These findings provide a foundational reference for the advancement of clinical strategies targeting the diagnosis and treatment of tauopathies.

Fig. 1 (a) TPE-P9 acts as an ‘off–on’ probe for Tau fibril detection. (b) ThT acts as an ‘always-on’ probe for Tau fibril detection. (c) TPE-P9 acts as a ‘steric zipper blocker’ to suppress hyperphosphorylation-induced Tau aggregation, thus diminishing pathological Tau production. (d) TPE-P9 acts as a ‘cross-seed’ to accelerate pathological Tau fibrillization, thus inhibiting toxic species internalization.

Results and discussion

Design and synthesis of TPE-P9

In an attempt to enhance the structural flexibility of TPE–NH₂, a maleimide group was incorporated. This led to the synthesis of TPE-Mal via a streamlined two-step reaction, with the efficiency surpassing 75%. The compound was characterized with ¹H-NMR, ¹³C-NMR, and high-resolution mass spectroscopy (HRMS) spectra. The general synthetic route and associated spectral data are depicted in Fig. 2a and Fig. S1–S3 (SI).

Fig. 2 (a) Synthetic route to TPE-Mal. (b) Lipophilic properties of the compounds (the values of c Log D represent the corresponding values of Log P at physiological pH 7.4, predicted by the online ADMETlab 2.0 platform).

TPE-P9 was prepared through a Michael addition reaction occurring between the luminescent unit (TPE-Mal) and the targeting moiety (CKVQIINKK, P9) under mild conditions. The selection of P9, a structurally modified peptide derived from the hexapeptide motif of Tau with the sequence VQIINK (P6), was inspired by a recent report demonstrating that the dissociation constant (K_d) of P6 for binding aggregation-prone Tau was over 17-fold lower than that for normal Tau, underscoring the potential of P6 as a highly selective targeting peptide for Tau aggregates.³⁰ To enhance conjugation, a cysteine residue was introduced at the N-terminus of P6 to serve as a linker between the peptide and the AIEgen. Considering that high BBB permeability is contingent upon optimal lipophilicity of the molecule,³¹ two lysine residues were incorporated to meticulously balance the hydrophobic and hydrophilic domains. As expected, this modification successfully conferred the synthesized TPE-P9 with suitable lipophilicity, as indicated by a calculated Log P value (c Log P) of 0.969 (Fig. 2b). The molecule was characterized by 1D (¹H, ¹³C, DEPT-135, DEPT-90) NMR, 2D (¹H–¹H COSY, HSQC, HMBC) NMR, and HRMS, with the relevant spectra, corresponding chemical shift assignments for the hydrogen and carbon nuclei, as well as plausible fragmentation pathways shown in Fig. S4–S6 (SI).

In some cases, condensation reactions between amion groups and carboxyl groups were employed to couple peptides with AIEgens.²⁰ However, this approach encountered various challenges, including the necessity for catalysts, multiple reactive sites, and intricate post-reaction processing. To avoid these issues, we adopted the Michael addition reaction as an alternative strategy for peptide–AIEgen conjugation. The Michael addition reaction is a pivotal organic transformation, in which a nucleophile selectively attacks the β-carbon of an α,β-unsaturated system to form a new carbon–carbon bond. Malemide derivatives, as highly reactive Micheal acceptors, are commonly utilized in conjugation with thiol-bearing molecules. This approach exhibits several distinct advantages, such as catalyst-free conditions, rapid reaction kinetics, near-quantitative conversion and remarkable specificity, thereby significantly reducing synthesis costs while optimizing yield.

Photophysical properties of the probes

The photophysical properties of TPE-P9 were evaluated in a binary solvent system of H₂O/DMF with incremental water fractions, where DMF acted as a good solvent and water as a poor solvent. As illustrated in Fig. 3a and c, TPE-P9 emitted minimal fluorescence in pure DMF due to high molecular dispersity. However, upon increasing the water content, aggregation-induced emission (AIE) was observed, with blue fluorescence detectable at the water fraction of 70%. The phenomenon was analogous to that of the unmodified TPE (Fig. S7, SI). The similarity in the AIE response between TPE-P9 and TPE demonstrated the minimal impact of P9 modification on the luminescent characteristics of TPE, supporting the rationality of selecting P9 for functionalization.

Fig. 3 The photophysical properties of TPE-P9 (10 µM) and ThT (10 µM). (a) Emission spectra of TPE-P9 in H₂O-DMF mixtures with varying water fractions (f_w, vol%), λ_ex = 330 nm. (b) Emission spectra of ThT in THF-DMF mixtures with varying THF fractions (f_t, vol%), λ_ex = 385 nm. (c) Fluorescence intensity changes with poor solvent fractions (f_p, vol%), where I₀ represents the maximum intensities in pure DMF for TPE-P9 and in pure THF for ThT, respectively. Inset: Enlarged ThT plot. (d) Photostability of ThT and TPE-P9 upon continuous light irradiation. I₀ and I are the maximum intensities of each dye recorded before and after exposure. Data are means ± SD (n = 3).

In a 99% water solution, the fluorescence intensity of TPE-P9 was 20-fold lower than that of TPE at equivalent concentrations, indicating the improved dispersibility of the modified AIE-active molecule in aqueous solution, addressing the prevalent limitation of traditional AIEgens being primarily applied in organic solvents.

In stark contrast to the AIE behavior observed for TPE-P9, the fluorescence intensities of ThT, a common dye for amyloid protein detection, showed an inverse relationship with increasing fractions of the poor solvent (Fig. 3b and c), a characteristic sign of the aggregation-caused quenching (ACQ) effect.³² This effect poses a risk of inaccurate detection when analyte concentrations are sufficiently high to cause excessive aggregation of the probe. Additionally, compared with TPE-P9, ThT exhibited markedly lower photostability under continuous light irradiation (Fig. 3d). These limitations, including ACQ susceptibility and poor photostability, might severely impede the practical applications of ThT.

From a chemical structure perspective, unlike the large planar structure of ThT, TPE-P9 features a central ethylene bond linking four phenyl rings with significant steric hindrance, creating a twisted propeller-like structure. In good solvents, the molecule exhibits rapid intramolecular motion, leading to weak or negligible emissions due to high nonradiative decay (k_nr). However, in solid states or when aggregated in poor solvents, its twisted structure restricts intramolecular motion and intermolecular π–π stacking, suppressing k_nr. This restriction allows intense emissions, as the radiative decay rate (k_r) could effectively compete with k_nr.³³

AIE-induced specific detection of Tau fibrils by TPE-P9 in aqueous solution

In a normal physiological state, Tau exhibits weak spontaneous nucleation. However, in tauopathies, Tau dissociates from microtubules and self-assembles into toxic species, causing neuronal dysfunction. During the simulation of fibrillar Tau in vitro, heparin was used as an inducer. Additionally, dithiothreitol (DTT), a reducing agent, was applied to disrupt intra- or inter-molecular disulfide bonds, facilitating the formation of more extended Tau structures for thorough interaction with the inducer.³⁴ The aggregation degrees of Tau under different inducing conditions were evaluated by ThT emitted intensities, as shown in Fig. S8 (SI). The net changes of ThT signals in systems with Tau, heparin, and DTT increased by over 9-fold, 3.7-fold, and 2-fold compared to systems with Tau alone, Tau and DTT, or Tau and heparin, respectively, highlighting the synergistic contributions of both the inducer and the antioxidant to Tau fibril formation in vitro. The prepared Tau fibrils were confirmed by TEM imaging, which revealed distinct fibrillar morphologies (Fig. 4a).

Fig. 4 (a) Time-dependent ThT (10 µM) fluorescence curves to monitor the aggregation kinetics of Tau protein with or without TPE-P9 (1 µM). Inset: TEM characterization with negative staining of the prepared fibrillar Tau. Scale bar, 200 nm. (b) Fluorescent spectra of TPE-P9 (1 µM) in the presence of different concentrations of Tau fibrils (0–4 µM) and its corresponding linear regression curve. (c) Fluorescent spectra of TPE-P9 (1 µM) in the absence and presence of Tau, Aβ fibrils and BSA (4 µM). (d) Fitted sigmoidal binding curve of the concentration-jump MST signal for the interaction between pathological Tau or normal Tau and TPE-P9 and for the interaction between pathological Tau and ThT. (e) Comparison of the selectivity of TPE-P9 (1 µM) and ThT (1 µM) towards 4 µM of Tau, Aβ fibrils and BSA. Data are means ± SD (n = 3). * The curves of BSA and TPE-P9 overlapped due to similar emission properties.

To minimize background fluorescence interference and eliminate false-positive signals arising from TPE-P9 self-aggregation, the critical concentration of TPE-P9 in the working aqueous environment was determined. This concentration served as a threshold below which negligible emissions were detectable. Detailed analysis (Fig. S9, SI) revealed that TPE-P9 exhibited a distinct concentration-dependent fluorescence response. Specifically, the luminogen produced weak fluorescence at concentrations ≤1 µM in a Tris-HCl buffer. On the other hand, aggregation kinetics monitored by ThT during co-incubation of Tau with TPE-P9 at 1 µM closely aligned with those of the pure protein (Fig. 4a), conforming that TPE-P9 at a low concentration below 1 µM did not perturb the natural aggregation process of Tau. Conversely, TPE-P9 at concentrations of 2–10 µM emitted intense fluorescence at 465 nm, with maximum intensities ranging from 48 to 400 a.u. Consequently, 1 µM was selected as the optimal working concentration of TPE-P9 for subsequent amyloid protein detection in this study, unless otherwise specified. Additionally, the probe's ideal incubation time was determined to be 10 minutes at room temperature (Fig. S10, SI), ensuring robust and reproducible results.

Following the assembly of fibrillar Tau in vitro and optimization of the detecting parameters, a series of spectroscopic tests were conducted to study the binding-induced emission of TPE-P9 when interacting with Tau fibrils. As depicted in Fig. 4b, TPE-P9-induced fluorescence was enhanced upon the introduction to Tau fibrils at various probe-to-analyte molar ratios of 1 : 0.2 to 1 : 4, and fluorescence intensities increased linearly from 20 to 54 a.u. with the correlation coefficient of 0.9951 at the adsorption peak of 465 nm. In contrast, with 4 µM Tau monomers, the highest emission at 465 nm was observed at 20 a.u., comparable to the signals from fibrils at a detected concentration of 0.2 µM (Fig. 4c). Furthermore, the binding affinity of TPE-P9 for fibrillar Tau (K_d = 4.46 µM) was higher than that for monomeric Tau (K_d = 71.3 µM) (Fig. 4d), signifying a preferential binding preference of TPE-P9 to fibrillar Tau over monomers, and the distinct emission intensities of the probe were closely associated with the aggregation state of the target.

Given the structure homology of hydrophobic β-sheet conformations among common fibrillar amyloid species, developing a probe capable of discerning Tau fibrils with high specificity remains a formidable challenge.¹⁹ Aβ fibrils, for instance, generally coexist with Tau fibrils in a pathological microenvironment, may interfere the response signals of TPE-P9. Therefore, the selectivity of TPE-P9 for Tau fibrils was systematically assessed. As shown in Fig. 4c and e, unlike ThT, TPE-P9 emitted weakly when introduced to Aβ fibrils, with the highest emission peak at 15 a.u. at a concentration of 4 µM, which was lower than the signals produced by Tau fibrils at a detected concentration of 0.2 µM. Meanwhile, in an aqueous solution containing 4 µM bovine serum albumin (BSA), the emission remained weak to negligible, and the curve just overlapped with that of the pure TPE-P9 solution. All of the observed phenomena indicate that TPE-P9 can recognize and bind to Tau aggregates in a sequence-dependent manner. Furthermore, TPE-P9 retained robust physicochemical stability in a thiol-rich environment, with fluorescence intensity fluctuations remaining within 20% over a 12-hour period, and the presence of inorganic ions or small molecular amino acids was observed to cause negligible interference with the dye emission properties (Fig. S11, SI). These merits enabled TPE-P9 to be a powerful and reliable tool for the specific detection of Tau fibrils in complex biological matrices.

Fluorescence imaging of pathological Tau by TPE-P9 in living cells

Motivated by the robust Tau-targeting efficacy of TPE-P9 in aqueous solution, we further explored its potential for imaging endogenous pathological Tau species in living HT22 neurons. In order to establish the cellular model of tauopathy, we utilized okadaic acid (OA), a protein phosphatase inhibitor, to disrupt the dynamic balance between Tau phosphorylation and dephosphorylation, thereby inducing an upregulation of intracellular pathological Tau.^25,35 Western blot (WB) analysis was conducted to validate the elevation of hyperphosphorylated Tau (p-Tau) levels following OA treatment. As depicted in Fig. S12 (SI), OA treatment resulted in a significant increase in p-Tau levels within HT22 cells, with the ratio of p-Tau to total Tau rising by 57.3%, confirming the successful establishment of a pathological Tau model. The in situ specific biosensing capability of the probe was evaluated using an inverted fluorescence microscope. As shown in Fig. 5a and b, the emission of TPE-P9 in the control group lacking OA was minimal, while a marked fluorescence enhancement was observed in OA-treated cells incubated with TPE-P9, which was consistent with ThT staining results. Notably, leveraging the pronounced off–on fluorescence properties of TPE-P9, background interference in detection was dramatically reduced compared to ThT, as evidenced by the more pronounced increase in fluorescence intensity in the treated group relative to the control. To further validate the reliability of TPE-P9 staining, co-labeling experiments were performed using TPE-P9 and an anti-p-Tau antibody. As shown in Fig. 5c, the labeling pattern of TPE-P9 was highly consistent with that of the p-Tau antibody, confirming the specificity and reliability of TPE-P9 for detecting pathological Tau. Due to the single-epitope specificity of antibodies, the selected anti-p-Tau antibody (targeting S396) exhibited intense fluorescence predominantly in regions with high expression of p-Tau S396. However, OA treatment induces phosphorylation at multiple sites across the Tau molecule, not limited to S396. In contrast, TPE-P9 binds to various pathological Tau species regardless of their specific phosphorylation sites, including but not limited to p-Tau S396. Consequently, the fluorescent signal of TPE-P9 covered a broader subcellular area, reflecting its ability to comprehensively detect pathological Tau variants. These findings collectively demonstrated that TPE-P9 held promise as a potential biosensor for in situ discrimination between normal and pathological cells in the field of tauopathies.

Fig. 5 (a) Fluorescence images and (b) relative intensity analysis of HT22 after incubation with TPE-P9 and ThT. Scale bars, 50 µm. Data are means ± SD (n = 3). (c) Co-localization imaging of TPE-P9 with anti-p-Tau antibody in HT22. Scale bars, 20 µm.

Mitigation of pathological Tau-mediated cytotoxicity via aggregation suppression

The hexapeptide P6, situated at the N-terminus of the repeat 2 (R2) domain and functioning as a steric zipper, has been identified as a critical mediator for both Tau assembly and intercellular propagation. Given that TPE-P9 is a derivative of P6, we hypothesized that its binding to the P6 motif could interfere with this key structural element, thereby modulating Tau aggregation and subsequent cytotoxicity. To confirm this hypothesis, we exploited the CCK-8 assay to evaluate the therapeutic effect of TPE-P9 on pathological Tau-induced apoptosis. As a control, TPE-P9 exhibited minimal toxicity to HT22 cells within the concentration range of 1–10 µM, with cell viabilities remaining above 90%. However, a reduction in cell viability to 83.8% was observed at a higher concentration of 20 µM. To explore the origin of this cytotoxicity, cells were separately treated with TPE-Mal and P9 at varying concentrations ranging from 1 to 20 µM. P9 demonstrated negligible toxicity with cell viabilities ranging from 94.9% to 108%, while TPE-Mal showed dose-dependent toxicity, with cell viabilities decreasing from 89.5% to 67.0% as its concentration increased (Fig. 6a). These results suggested that the cell toxicity observed with TPE-P9 was predominantly attributed to the TPE-Mal moiety, further corroborating the critical necessity of P9 modification. This structural refinement not only augmented the molecular targeting precision but also ameliorated the biocompatibility of the AIE core.

Fig. 6 (a) TPE-P9, P9, and TPE-Mal cytotoxicity assessment at varying concentrations of 1–20 µM. (b) Dose-dependent HT22 cell protection of TPE-P9 on OA and GSK3β-induced Tau oligomers using the CCK-8 assay. (c) Western blot analysis of p-Tau and Tau monomers after Tau phosphorylation in the presence of GSK3β at various times. (d) Representative AFM images of Tau and phosphorylation-induced Tau aggregates in the absence of or in the presence of TPE-P9 after 72 h of co-incubation. Scale bars, 100 nm. (e) Western blot analysis of p-Tau and Tau monomers after GSK3β-induced Tau phosphorylation for 72 h in the absence (−) or presence of TPE-P9 at different molar ratios. +0.1 : 1, ++1 : 1, +++5 : 1, ++++10 : 1 (TPE-P9 : Tau). Data are means ± SD (n = 3, 5). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (one-way ANOVA). (f) Apoptosis assay of HT22 cells co-incubated with 5 µM of either native Tau, p-Tau (without TPE-P9), or TPE-P9 pretreated p-Tau (TPE-P9–p-Tau) for 48 h. The quadrants of R1, R2, R3 and R4 respectively represent the dead, late apoptotic, early apoptotic, and live cells.

As previously mentioned, post-translational modifications (PTMs) of proteins play a crucial role in regulating Tau's biological activity. Under malfunctioning conditions, Tau protein generally undergoes abnormal phosphorylation at numerous sites, leading to global charge redistribution and conformation transition. These changes drive Tau self-assembly into pathological aggregates, potentially triggering various diseases.³⁶ To investigate the neuroprotective potential of TPE-P9 against aberrant PTMs-induced apoptosis, OA was applied to perturb the physiological PTM process. Treatment of HT22 cells with 100 nM OA significantly increased apoptosis, resulting in a 26.6% decrease in cell viability. However, with the addition of TPE-P9, cell viability gradually increased within the range of 1 to 10 µM. When the concentration of TPE-P9 reached 5 µM and 10 µM, it completely reversed the cytotoxicity induced by OA, with the cell survival rates increased by 21.5% and 19.2%, respectively. Interestingly, at a concentration of 20 µM, TPE-P9 exhibited weaker therapeutic effects, increasing viability by only 8.26%, likely due to inherent cell toxicity at this concentration (Fig. 6b).

Mounting literature studies have pointed out that cell toxicity primarily stemmed from Tau oligomers, rather than monomers.^34,37 Building upon this understanding, we postulated that TPE-P9 protected neurons from pathological Tau-caused cytotoxicity by intervening with its aggregation cascade. To validate this conjecture, we performed WB assays to assess the impact of TPE-P9 on aggregation levels of p-Tau, which spontaneously assembled into toxic species. For the in vitro preparation of p-Tau, glycogen synthase kinase-3β (GSK3β) was employed as the inducer, as the resulting phosphorylation patterns on Tau closely resemble those identified in paired helical filaments of AD.³⁸ The WB images of p-Tau were obtained before and after the treatment with GSK3β. In the control group lacking GSK3β, negligible p-Tau protein was observed. The content of p-Tau increased with incubation time from 4 hours to 24 hours, reaching equilibrium within another 24 hours. Subsequently, p-Tau began to self-assemble, as evidenced by the significant decrease in both p-Tau and total Tau monomers (Fig. 6c). This self-aggregation process was further confirmed by atomic force microscopy (AFM) images, which revealed the formation of highly polymorphic oligomers over a 72-hour period. The average height and length of these oligomers were 2.52 nm and 22.82 nm, respectively, corresponding to 219% and 119% of the values observed for Tau, respectively (Fig. 6d). However, the presence of TPE-P9 suppressed p-Tau aggregation after 72 hours of co-incubation, as reflected by increased relative levels of both p-Tau and total Tau monomers. Specifically, TPE-P9 significantly increased the relative levels of monomers by 81.7% and 108% for p-Tau, and 74.8% and 136% for total Tau at molar ratios of 1 : 1 and 5 : 1 (TPE-P9 : Tau), respectively, compared to the group without TPE-P9 (Fig. 6e). The aggregation-inhibitory capacity of TPE-P9 was also corroborated by AFM analysis, with a reduction in the average height and length by 38.1% and 16.4% for the p-Tau sample, respectively, bringing them closer to the size observed for the normal Tau sample (Fig. 6d). The prepared p-Tau with various aggregation statuses was introduced to HT22 cells. As expected, p-Tau demonstrated increasing cytotoxicity over 72 hours of incubation, while unmodified Tau had minimal effects (Fig. S13, SI). Pre-incubation with TPE-P9 significantly alleviated the cytotoxicity of p-Tau aggregates, with cell viability increased by 9.49–35.4% observed at TPE-P9 : Tau ratios of 0.2 : 1, 1 : 1, 2 : 1, and 4 : 1 (Fig. 6b). This trend in cell viability was in excellent agreement with the results from flow cytometry (FACS) analysis. Specifically, compared to the p-Tau-only treatment group, the TPE-P9-pretreated p-Tau group (TPE-P9–p-Tau group) exhibited a 38.6% increase in the proportion of viable cells, as quantified by FACS (Fig. 6f). The above findings demonstrated that TPE-P9 could suppress the self-aggregation process of p-Tau, thus preventing the formation of toxic Tau oligomers induced by hyperphosphorylation and alleviating associated cytotoxicity.

To elucidate the modulation mechanisms in greater depth, we propose that TPE-P9, an analog of P6, is able to bind to the P6 motif of p-Tau and acts as a steric zipper blocker to mask the aggregation-driven core region. Additionally, the hydrophobic microdomains created by the AIE-active luminescent unit facilitate the sequestration of bound p-Tau on the surface, promoting steric repulsion and limiting further intermolecular interactions, thereby preventing higher-order aggregation of p-Tau. However, with the continuous increase of intervening concentration to 10 : 1 (molar ratio of TPE-P9 : Tau), the inhibition effect of TPE-P9 decreased, as reflected by the increase of p-Tau and total Tau monomers by 50.6% and 55.4%, respectively. To further analyze the reasons, we believe that the presence of excessive TPE-P9 may self-assemble into a cross-seed (Fig. S14, SI), which could drive p-Tau aggregation and partially offset the suppression effects of TPE-P9. Extensive biochemical experiments are necessary to delve into the details of the interactions between p-Tau and TPE-P9 at the molecular level.

Alleviation of pathological Tau-induced cytotoxicity by spread inhibition

In a pathological state, aberrant PTMs convert normal Tau into pathological species. These toxic conformers are secreted from neurons through various mechanisms, including passive diffusion and exocytosis.^39,40 Exogenous misfolded proteins are then taken up by neighboring cells via extracellular vesicles, through tunneling nanotube formation, or in a free molecular form.⁴¹ While the precise molecular mechanisms underlying these processes remain under active debate, current evidence suggests that Tau internalization typically relies on active endocytic pathways or micropinocytosis. The routes including those mediated by heparan sulfate proteoglycans (HSPGs), lipoprotein receptor-related protein 1 (LRP1),^42,43 the presynaptic scaffolding protein bassoon,⁵ clathrin, and caveolae⁴⁴ have been implicated in the internalization of pathological Tau, leading to the cell-to-cell spread of tauopathy.⁴⁵ Using cryo-electron tomography to reconstruct molecular-resolution tomograms, researchers have shown that the released Tau primarily exists as filamentous structures, including paired helical filaments (PHFs) and straight filaments (SFs).³⁹ Early studies revealed that p-Tau alone was insufficient to form filaments in vitro, likely due to the lack of β-sheet structures. This gap was addressed when sulphated glycosaminoglycans (sGAGs) are present. Goedert and his group members’ work showed that sGAGs promoted Tau filament formation in a phosphorylation-independent manner, yielding structures highly similar to those observed in the AD brain.^46,47 Building on this, heparin, a potent sGAG, was applied to drive the self-aggregation of inert Tau by exposing the aggregation-prone VQIINK and VQIVYK motifs,⁴⁸ thereby generating in vitro Tau seed mimics.

Treating HT22 cells with heparin-induced Tau seeds led to 27.0% decrease in cell viability. In the case of the Tau–TPE-P9 system, after 48 hours of incubation of TPE-P9 with toxic Tau seeds at TPE-P9 : Tau ratios of 0.2 : 1, 1 : 1, 2 : 1, and 4 : 1, cell viability increased by 2.71%, 3.24%, 3.47%, and 9.27%, respectively, compared to the group without TPE-P9 (Fig. 7a). Flow cytometry assays were performed to assess the impact of TPE-P9 on Tau seeds propagation. As shown in Fig. 7b, TPE-P9 treatment reduced the internalization of heparin-induced Tau aggregates in a dose-dependent manner, as revealed by a significant decrease in the median fluorescence intensity (MFI) by 11%, 13%, 23%, and 33% at TPE-P9 : Tau ratios even as low as 0.2 : 1, 0.3 : 1, 0.4 : 1, and 0.8 : 1, respectively.

Fig. 7 (a) Dose-dependent HT22 cell protection of TPE-P9 on heparin-induced pathological Tau seeds using the CCK-8 assay. (b) Dose-dependent inhibitory effect of TPE-P9 on the cellular internalization of pathological Tau. (c) Western blot analysis of Tau monomers after heparin induced aggregation in the absence (−) or presence of TPE-P9 at different molar ratios. (d) Representative AFM images of heparin-induced Tau aggregates in the absence of or in the presence of TPE-P9 after 72 h of co-incubation. Scale bars, 100 nm. +0.1 : 1, ++1 : 1, +++5 : 1, ++++10 : 1 (TPE-P9 : Tau). Data are means ± SD (n = 3, 6). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (one-way ANOVA).

A series of studies have pointed out that the internalization of Tau seeds is contingent upon both conformational states and polymeric dimensions. We therefore proposed that TPE-P9 exerted its protective effect by modulating the formation of Tau seeds. To validate this hypothesis, WB assays instead of ThT staining were conducted to evaluate the impact of TPE-P9 on heparin-induced Tau aggregation. This methodological choice was rooted in the observation that TPE-P9 exhibited a higher binding affinity for pathological Tau (K_d = 4.46 µM) compared to ThT (K_d = 9.26 µM) (Fig. 4d), signifying potential competitive binding when both compounds simultaneously interacted with Tau. Such competition could result in fluorescence decay of ThT, compromising the accuracy of ThT aggregation monitoring assays. As shown in Fig. 7c, when TPE-P9 co-existed with Tau at different molar ratios of 0.1 : 1, 1 : 1, 5 : 1, and 10 : 1 (TPE-P9 : Tau) with a fixed Tau concentration of 40 µM, TPE-P9 accelerated Tau aggregation, as indicated by decreased relative levels of Tau monomers. Specifically, TPE-P9 significantly reduced the relative levels of Tau monomers by 31.6% and 29.2% at molar ratios of 1 : 1 and 5 : 1 (TPE-P9 : Tau), respectively. AFM analysis revealed that the addition of TPE-P9 resulted in the formation of thicker and longer amyloid aggregates, with average height and length increases of 84.6% and 103%, respectively, compared to the amyloid controls lacking TPE-P9 (Fig. 7d). These AFM observations were in excellent agreement with the WB results, providing converging evidence for TPE-P9's intervening effect on heparin-induced Tau aggregation. These above experimental results demonstrated that the protective effect of TPE-P9 against heparin-induced Tau seeds’ cytotoxicity is primarily attributed to its role in accelerating Tau fibrillogenesis, allowing it to swiftly circumvent the most detrimental oligomeric forms^20,49 and diminishing the uptake of pathological Tau by cells.

Numerous strategies have been developed to mitigate pathological Tau-induced cytotoxicity, such as inhibiting Tau aggregation, modulating Tau phosphorylation, restraining exosome releasing, and suppressing Tau endocytosis.^50,51 However, most of these strategies are limited in that they only target either reducing pathological Tau production or halting its spread (Table S1, SI), resulting in poor therapeutic effects of tauopathies. In this study, we intervened in both the process of pathological Tau oligomerization and its propagation among cells, significantly enhancing the survival rate of neurons. The mechanisms of action are illustrated in Fig. 8.

Fig. 8 General mechanisms of tauopathy therapeutics focus on inhibiting Tau aggregation and halting its spreading, thereby preventing neuron death. More details here are listed in Table S1 (SI).

Conclusions

In this study, a novel dual-functional probe, TPE-P9, was designed and synthesized through a Michael addition reaction between the thiol-modified peptide (P9) and the maleimide moiety of TPE-Mal. Modification with P9, while preserving the AIE properties, not only enables the probe to selectively distinguish pathological Tau at both aqueous and living cell levels, but also optimizes the hydrophobic–hydrophilic balance and enhances the biocompatibility of the luminescent part. The appropriate lipophilicity empowers TPE-P9 to potentially overcome the permeation challenge of the BBB. Compared to most commercial dyes such as ThT, the synthesized TPE-P9 exhibited significantly better photostability, higher detection specificity, stronger affinity to the analytes, and lower background fluorescence interference. Importantly, besides its capability for amyloid detection, TPE-P9 also demonstrated powerful modulation of Tau amyloid aggregation, underscoring its promising applications in both bioimaging and biotherapy. The comparison between TPE-P9 and other amyloid protein detectors or modulators is summarized in Fig. 9.

Fig. 9 Comparison of various amyloid protein detectors and modulators in terms of their specificities, K_d values and relative modulation rates. More details here are listed in Table S2 (SI).

As an amyloid detector, TPE-P9 demonstrated its ability to detect conformation-specific and sequence-dependent Tau aggregates in aqueous solution and living neurons, exhibiting turn-on fluorescence. Additionally, as an amyloid modulator, TPE-P9 intervenes in the progression of tauopathies from two distinct aspects. In the case of hyperphosphorylation-induced pathological Tau, TPE-P9 suppressed p-Tau self-assembly by both shielding the aggregation core of p-Tau and concurrently increasing the steric hindrance of intermolecular interactions, reducing toxic oligomer production, and protecting cells from p-Tau-induced apoptosis. This resulted in increased cell viability by 35.4%. As for the released cytotoxic Tau seeds, TPE-P9 showed general cross-seeding ability, accelerating fibrillar formation and bypassing toxic species. This reduced endocytosis of toxic species and blocked their ‘prion-like’ spread among neighboring cells, thereby increasing the viability of pathological Tau-treated HT22 cells by 9.27%. This research addresses the urgent need for a novel strategy to achieve simultaneous real-time visualization and therapeutics for early tauopathies.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data that support the findings of this study are available in the supplementary information (SI) of this article. Supplementary information: materials, instrumentations, experimental methods, NMR (1D & 2D), HRMS, fluorescence, western-blot, viability of HT22, comparison of TPE-P9 with various tauopathy therapeutics, amyloid protein detectors and modulators (PDF). See DOI: https://doi.org/10.1039/d5tb01783a.

References

1. J. N. Rauch, S. H. Olson and J. E. Gestwicki, Cold Spring Harbor Perspect. Med., 2017, 7, a024034.
2. M. Goedert, C. M. Wischik, R. A. Crowther, J. E. Walker and A. Klug, Proc. Natl. Acad. Sci. U. S. A., 1988, 85, 4051–4055.
3. D. P. Hanger, J. C. Betts, T. L. F. Loviny, W. P. Blackstock and B. H. Anderton, J. Neurochem., 2002, 71, 2465–2476.
4. M. Morris, G. M. Knudsen, S. Maeda, J. C. Trinidad, A. Ioanoviciu, A. L. Burlingame and L. Mucke, Nat. Neurosci., 2015, 18, 1183–1189.
5. J. Ye, H. Wan, S. Chen and G. P. Liu, Neural. Regener Res., 2024, 19, 1489–1498.
6. A. Tarutani, T. Adachi, H. Akatsu, Y. Hashizume, K. Hasegawa, Y. Saito, A. C. Robinson, D. M. A. Mann, M. Yoshida, S. Murayama and M. Hasegawa, Acta Neuropathol., 2022, 143, 613–640.
7. S. Dhillon, Drugs, 2021, 81, 1437–1443.
8. S. M. Hoy, Drugs, 2023, 83, 359–365.
9. C. Kang, Drugs, 2024, 84, 1313–1318.
10. V. Lindström, T. Fagerqvist, E. Nordström, F. Eriksson, A. Lord, S. Tucker, J. Andersson, M. Johannesson, H. Schell, P. J. Kahle, C. Möller, P. Gellerfors, J. Bergström, L. Lannfelt and M. Ingelsson, Neurobiol. Dis., 2014, 69, 134–143.
11. M. J. Koper, S. Moonen, A. Ronisz, S. Ospitalieri, Z. Callaerts-Vegh, D. T’Syen, S. Rabe, M. Staufenbiel, B. D. Strooper, S. Balusu and D. R. Thal, Sci. Transl. Med., 2024, 16, eadf5128.
12. F. Lopera, C. Marino, A. S. Chandrahas, M. O’Hare, N. D. Villalba-Moreno, D. Aguillon, A. Baena, J. S. Sanchez, C. Vila-Castelar, L. Ramirez Gomez, N. Chmielewska, G. M. Oliveira, J. L. Littau, K. Hartmann, K. Park, S. Krasemann, M. Glatzel, D. Schoemaker, L. Gonzalez-Buendia, S. Delgado-Tirado, S. Arevalo-Alquichire, K. L. Saez-Torres, D. Amarnani, L. A. Kim, R. C. Mazzarino, H. Gordon, Y. Bocanegra, A. Villegas, X. Gai, M. Bootwalla, J. Ji, L. Shen, K. S. Kosik, Y. Su, Y. Chen, A. Schultz, R. A. Sperling, K. Johnson, E. M. Reiman, D. Sepulveda-Falla, J. F. Arboleda-Velasquez and Y. T. Quiroz, Nat. Med., 2023, 29, 1243–1252.
13. P. Giannakopoulos, F. R. Herrmann, T. Bussière, C. Bouras, E. Kövari, D. P. Perl, J. H. Morrison, G. Gold and P. R. Hof, Neurology, 2003, 60, 1495–1500.
14. Y. Ishida, M. Inoue, T. Inoue, A. Ojida and I. Hamachi, Chem. Commun., 2009, 2848–2850.
15. X. Gao, L. Wang, H. L. Huang, L. L. Wang, J. L. Yao, S. Shi and T. M. Yao, Analyst, 2015, 140, 7513–7517.
16. C. Li, P. Chen, X. Ma, X. Lin, S. Xu, S. Niazi and Z. Wang, Nano Res., 2023, 16, 5391–5400.
17. X. He, T. Zeng, Z. Li, G. Wang and N. Ma, Angew. Chem., Int. Ed., 2015, 55, 3073–3076.
18. W. Zhang, X. Liu, P. Li, W. Zhang, H. Wang and B. Tang, Anal. Chem., 2020, 92, 3716–3721.
19. T. Zhang, X. Chen, C. Yuan, X. Pang, P. Shangguan, Y. Liu, L. Han, J. Sun, J. W. Y. Lam, Y. Liu, J. Wang, B. Shi and B. Z. Tang, Angew. Chem., Int. Ed., 2023, 62, e202211550.
20. Y. Tang, D. Zhang, X. Gong and J. Zheng, Adv. Funct. Mater., 2022, 32, 2208022–2208033.
21. J. Luo, Z. Xie, J. W. Y. Lam, L. Cheng, H. Chen, C. Qiu, H. S. Kwok, X. Zhan, Y. Liu, D. Zhu and B. Z. Tang, Chem. Commun., 2001, 1740–1741.
22. W. He, Y. Yang, Y. Qian, Z. Chen, Y. Zheng, W. Zhao, C. Yan, Z. Guo and S. Quan, Aggregate, 2024, 5, e412.
23. Q. He, M. Wang, L. Zhao, B. Xu, W. Tian and L. Zhang, Aggregate, 2024, 5, e505.
24. Y. Li, W. He, Q. Peng, L. Hou, J. He and K. Li, Food Chem., 2019, 287, 55–60.
25. L. Ji, X. Zhou, R. Liu, L. Wang, Y. Hu, J. Gu, Z. Li, C. Li, T. Huang and Y. Yu, Sens. Actuators, B, 2023, 380, 133415–133422.
26. J. Xu, L. Fang, M. Shi, Y. Huang, L. Yao, S. Zhao, L. Zhang and H. Liang, Chem. Commun., 2019, 55, 1651–1654.
27. S. Shao, Z. Li, H. Cheng, S. Wang, N. G. Perkins, P. Sarkar, W. Wei and M. Xue, J. Am. Chem. Soc., 2018, 140, 13586–13589.
28. Y. Cheng, A. E. Clark, J. Zhou, T. He, Y. Li, R. M. Borum, M. N. Creyer, M. Xu, Z. Jin, J. Zhou, W. Yim, Z. Wu, P. Fajtová, A. J. O’Donoghue, A. F. Carlin and J. V. Jokerst, ACS Nano, 2022, 16, 12305–12317.
29. Z. Zhang, M. Kang, H. Tan, N. Song, M. Li, P. Xiao, D. Yan, L. Zhang, D. Wang and B. Z. Tang, Chem. Soc. Rev., 2022, 51, 1983–2030.
30. L. Xu, Y. Ding, F. Ma, Y. Chen, G. Chen, L. Zhu, J. Long, R. Ma, Y. Liu, J. Liu, F. Huang and L. Shi, Nano Today, 2022, 43, 101388–101397.
31. Q. Liu, Z. Chen, A. Guiseppi-Elie, F. Meng and L. Luo, Smart Mol., 2025, 3, e20240052.
32. W. Fu, C. Yan, Z. Guo, J. Zhang, H. Zhang, H. Tian and W. H. Zhu, J. Am. Chem. Soc., 2019, 141, 3171–3177.
33. Y. Chen, J. W. Y. Lam, R. T. K. Kwok, B. Liu and B. Z. Tang, Mater. Horiz., 2019, 6, 428–433.
34. M. Liu, D. Sui, T. Dexheimer, S. Hovde, X. Deng, K. W. Wang, H. L. Lin, H. T. Chien, H. K. Kweon, N. S. Kuo, C. A. Ayoub, D. Jimenez-Harrison, P. C. Andrews, R. Kwok, D. A. Bochar, J. Kuret, J. Fortin, Y. G. Tsay and M. H. Kuo, Mol. Neurobiol., 2020, 57, 4704–4719.
35. L. Ge and Y. Tian, Anal. Chem., 2019, 91, 3294–3301.
36. I. Landrieu, L. Lacosse, A. Leroy, J. M. Wieruszeski, X. Trivelli, A. Sillen, N. Sibille, H. Schwalbe, K. Saxena, T. Langer and G. Lippens, J. Am. Chem. Soc., 2006, 128, 3575–3583.
37. K. Flach, I. Hilbrich, A. Schiffmann, U. Gärtner, M. Krüger, M. Leonhardt, H. Waschipky, L. Wick, T. Arendt and M. Holzer, J. Biol. Chem., 2012, 287, 43223–43233.
38. C. A. Rankin, Q. Sun and T. C. Gamblin, Mol. Neurodegener., 2007, 2, 12–25.
39. S. L. Fowler, T. S. Behr, E. Turkes, D. P. O’Brien, P. M. Cauhy, I. Rawlinson, M. Edmonds, M. S. Foiani, A. Schaler, G. Crowley, S. Bez, E. Ficulle, E. Tsefou, R. Fischer, B. Geary, P. Gaur, C. Miller, P. D’Acunzo, E. Levy, K. E. Duff and B. Ryskeldi-Falcon, Nat. Neurosci., 2025, 28, 40–48.
40. R. Das, A. A. Balmik and S. Chinnathambi, J. Neuroinflammation, 2020, 17, 10–24.
41. C. Danis, E. Dupré, T. Bouillet, M. Denéchaud, C. Lefebvre, M. Nguyen, J. Mortelecque, F.-X. Cantrelle, J.-C. Rain, X. Hanoulle, M. Colin, L. Buée and I. Landrieu, Nat. Commun., 2025, 16, 3162.
42. X. Tian, E. K. Rivera, I. Hwang, A. A. Nikitina, J. A. Smith, Y. J. Cho, J. Sala-Jarque, A. DuBose, C. Andriola, T. Gollan-Myers and K. S. Kosik, Alzheimer's Dementia, 2025, 21, e70589.
43. A. T. Marvian, T. Strauss, Q. Tang, B. J. Tuck, S. Keeling, D. Rüdiger, N. Mirzazadeh Dizaji, H. Mohammad-Beigi, B. Nuscher, P. Chakraborty, D. S. Sutherland, W. A. McEwan, T. Köglsperger, S. Zahler, M. Zweckstetter, S. F. Lichtenthaler, W. Wurst, S. Schwarz and G. Höglinger, Mol. Neurodegener., 2024, 19, 100.
44. N. Puangmalai, N. Bhatt, M. Montalbano, U. Sengupta, S. Gaikwad, F. Ventura, S. McAllen, A. Ellsworth, S. Garcia and R. Kayed, Cell Death Dis., 2020, 11, 314.
45. J. W. Wu, M. Herman, L. Liu, S. Simoes, C. M. Acker, H. Figueroa, J. I. Steinberg, M. Margittai, R. Kayed, C. Zurzolo, G. Di Paolo and K. E. Duff, J. Biol. Chem., 2013, 288, 1856–1870.
46. J. Mendieta, M. A. Fuertes, R. Kunjishapatham, I. Santa-María, F. J. Moreno, C. Alonso, F. Gago, V. Muñoz, J. Avila and F. Hernández, Biochim. Biophys. Acta, 2005, 1721, 16–26.
47. M. Goedert, R. Jakes, M. G. Spillantini, M. Hasegawa, M. J. Smith and R. A. Crowther, Nature, 1996, 383, 550–553.
48. H. Mirbaha, D. Chen, O. A. Morazova, K. M. Ruff, A. M. Sharma, X. Liu, M. Goodarzi, R. V. Pappu, D. W. Colby, H. Mirzaei, L. A. Joachimiak and M. I. Diamond, eLife, 2018, 7, 36584–36612.
49. P. Wang, F. L. Cascio, J. Gao, R. Kayed and X. Huang, Chem. Commun., 2018, 54, 10120–10123.
50. A. Dominguez-Meijide, E. Vasili and T. F. Outeiro, Brain Sci., 2020, 10, 858–886.
51. J. Götz, A. Ittner and L. M. Ittner, Br. J. Pharmacol., 2012, 165, 1246–1259.

---