Tau protein aggregation is a hallmark of neurodegenerative diseases, including Alzheimer’s disease, making the development of anti-aggregation therapeutics a critical area of research. Progress in drug discovery has been hindered by the lack of efficient screening methods that accurately reflect cellular conditions. We present a high-throughput cell-based assay utilizing split GFP technology to monitor tau aggregation in living cells. Our system employs suspension-adapted HEK293 cells co-transfected with tau proteins fused to complementary GFP fragments, producing fluorescent signals upon tau aggregation. Notably, our system demonstrates tau aggregation without external aggregation inducers, likely due to the enhanced protein expression in suspension-adapted cells. Validation with a known urea-based tau aggregation inhibitor showed dose-dependent reduction in fluorescence, corresponding to decreased tau aggregation. The assay’s flow cytometry compatibility enables rapid, quantitative analysis of large sample sets while allowing simultaneous assessment of compound efficacy and cytotoxicity. This method advances tau aggregation monitoring and drug discovery by providing a physiologically relevant platform for identifying novel anti-tau aggregation therapeutics.
Tauopathies, including Alzheimer’s disease (AD), are a class of neurodegenerative disorders characterized by the pathological aggregation of the microtubule-associated protein Tau into neurofibrillary tangles (NFTs). These aggregates disrupt neuronal function, contributing to synaptic loss, neuronal death, and cognitive decline. Therapeutic interventions targeting Tau aggregation represent a promising strategy for mitigating disease progression. However, identifying effective compounds to prevent or reverse Tau aggregation requires robust and scalable screening platforms capable of capturing the complexity of Tau pathology in a physiologically relevant manner.
Contradictory to its pathological aggregation, Tau is an intrinsically disordered and highly soluble protein under normal physiological conditions. Its transition to a pathogenic, aggregation-prone state involves a cascade of post-translational modifications (PTMs) such as hyperphosphorylation, acetylation, and truncation, as well as conformational changes within neurons. Mutations in the Tau gene (MAPT), imbalances in expression of 3R and 4R isoforms, and altered PTM profiles have been shown to tip the balance toward aggregation. These factors collectively lead to the formation of pathological Tau aggregates, including paired helical filaments (PHFs) and neurofibrillary tangles (NFTs), which are among the hallmarks of Tau-related neurodegeneration.
Multiple reviews addressing the cellular and pathological functions of tau and the events leading to disease are available.
In vitro studies of Tau aggregation are valuable for understanding the molecular structure of PHFs and have been widely used in drug screening efforts. However, progress has been limited by an incomplete understanding of Tau aggregation dynamics and the lack of reliable models that reflect the biological complexity of Tau pathology. Recombinant full length tau purified from Escherichia coli exhibits limited aggregation potential due to the absence of essential PTMs. To address this limitation, researchers have employed truncated Tau isoforms, such as K18 and K19 (also known as repeat domains, RD) which contain the microtubule-binding region and aggregate more readily than full-length Tau. Additionally, disease-relevant mutations, such as P301S, P301L and ΔK280, enhance β-sheet formation and have been incorporated into models to increase aggregation efficiency. The aggregation rate is further accelerated by the addition of artificial cofactors, which neutralize Tau’s lysine-rich positive charge. These cofactors fall into two main categories: polyanions (e.g., heparin, RNA, polyglutamate and fatty acids or fatty acid-like molecules (e.g., arachidonic acid, docosahexaenoic acid). Polyanions, such as heparin, are particularly effective at promoting the polymerization of truncated Tau fragments (e.g., K18 and K19), while arachidonic acid enhances the aggregation of full-length Tau more efficiently than heparin.
Current approaches for detecting Tau aggregation in vitro rely heavily on fluorescence probes like thioflavin to monitor β-sheet formation and structural analyses using transmission electron microscopy. While these methods provide valuable mechanistic insights, they are resource-intensive, time-consuming, and dependent on non-physiological starting materials. Although these approaches improve aggregation rates and provide insights into Tau pathology, they rely on artificial systems that lack the complexity of the native cellular environment. As such, these models may not fully replicate the physiological conditions of Tau aggregation and are often unsuitable for large-scale, high-throughput drug screening campaigns. Furthermore, they do not account for the effects of PTMs, co-factors, or intracellular interactions that are crucial to Tau aggregation.
In vivo models, such as transgenic mice, offer a closer representation of human biology but are limited by ethical concerns, high costs, and a lack of scalability. These models are not practical for large-scale drug discovery efforts. As an alternative, cell-based assays provide a physiologically relevant and scalable platform for studying Tau aggregation. Mammalian cells can introduce post-translational modifications (PTMs), such as phosphorylation, acetylation, and glycosylation, that are absent in bacterial expression systems and more closely resemble those found in human proteins compared to yeast or insect cell systems. This makes mammalian cells a more suitable platform for studying drug effects on proteins and a potential intermediate model for investigating Tau aggregation dynamics. Previous studies with HEK cells have shown that glycosylation patterns of various human proteins are similar to native types, and other studies using HEK cell lines suggest that Tau expressed in these systems exhibits phosphorylation patterns comparable to those found in neuronal contexts. Additionally, cell-based systems enable simultaneous assessment of compound cell penetration — which is particularly important for Tau-targeting drugs, as Tau is an intracellular protein — along with cell viability and aggregation efficiency, making them well-suited for high-throughput applications.
There have been several cell-based assays for Tau aggregation described in the literature, as comprehensively reviewed by Lim et al.. To provide context for the current work, we summarize these systems to highlight how our assay, presented in this paper, differs from and builds upon previous approaches. One of the earliest cell-based models for Tau aggregation was developed by Mandelkow’s group. They generated a Tau-inducible cell line to address the toxicity associated with Tau aggregation in cells. Using an inducible system regulated by doxycycline, they overexpressed the truncated K18 Tau isoform in N2a neuroblastoma cells. While this approach demonstrated robust aggregation detectable by thioflavin S (ThS) staining, aggregation was observed only in the ΔK280 mutant and not in wild-type K18 Tau, even after nine days of expression. This highlights the challenges of inducing aggregation in physiologically relevant isoforms without introducing strong aggregation-prone mutations. Kuret’s and Lee’s groups showed full-length Tau-40 (2N4R Tau) aggregation in various cell lines only after adding small-molecule agonist of Tau aggregation (Congo red) or preformed Tau fibrils. These models have significantly advanced the understanding of intracellular Tau aggregation and its associated cellular toxicity. However, most of these systems rely on secondary methods such as ThS staining or immunostaining against phosphorylated Tau to confirm aggregation. While effective, these approaches are labor-intensive and less suited for high-throughput screening applications.
To overcome these limitations, fluorescence-based systems have been developed, incorporating fluorescent tags such as GFP, CFP, or YFP fused to various Tau isoforms like Tau-40, K18, or K19, including mutated variants. These models enable real-time monitoring of intracellular Tau expression in living cells without requiring secondary detection methods. However, these systems often encounter challenges in quantifying specific Tau-Tau interactions, as all Tau proteins in the assay are fluorescent, making it difficult to distinguish between aggregated and non-aggregated forms.
To address this, fluorescence resonance energy transfer (FRET) and bimolecular fluorescence complementation (BiFC) assays have been widely employed to study protein-protein interactions in general, as extensively reviewed elsewhere, and have also been adapted for Tau-Tau interactions. For instance, Johnson’s group utilized FRET by tagging full-length Tau with CFP and YFP and expressing them in HEK293 cells, enabling the detection of GSK3β-induced aggregation via energy transfer when the proteins were in close proximity. Similarly, Diamond’s group applied FRET to study the aggregation of the K18 version of Tau and explored its trans-cellular propagation. Their findings demonstrated that extracellular Tau fibrils could be taken up by neighboring cells, inducing intracellular aggregation and illustrating FRET’s utility in tracking both aggregation and propagation in living cells.
In addition, they created a stable HEK293T cell line expressing the K18 Tau P301S mutant tagged with CFP and YFP, in which Tau aggregation can be induced by fibrils or other compounds. This cell line, available from ATCC (CRL-3275), can also be used in seeding assays that directly translate to human brain homogenates, further emphasizing the utility of FRET in studying Tau pathology.
While FRET provides valuable insights into Tau aggregation, it has notable limitations, including the need for specialized equipment, low sensitivity, and potential interference from large fluorescent tags. Additionally, its efficiency depends on spectral overlap and dipole orientation of the fluorophores, which can complicate analysis. FRET also suffers from a limited dynamic range, weaker signals without amplification, and challenges in detecting stable protein-protein interactions. These drawbacks highlight the need for complementary methods, such as BiFC, to overcome these limitations.
Johnson’s group utilized a split GFP complementation technique as a “turn-off” sensor to quantify Tau aggregation in cells. In this approach, full-length Tau was fused to a small GFP fragment (GFP11) and co-expressed with a larger GFP fragment (GFP1-10). Fluorescence was generated when Tau remained as monomers, but aggregation prevented GFP reconstitution. While effective in demonstrating that mutants prone to aggregation do not produce fluorescence, this “turn-off” system is limited in its ability to monitor early aggregation events, such as soluble intermediates. Additionally, its application in screening campaigns for aggregation inhibitors or reversers is constrained by the extremely long half-life of the N-GFP and C-GFP interaction, estimated at 10 years. The Venus-based BiFC system, using split Venus fragments (VN173 and VC155) to label Tau, addresses this limitation by acting as a “turn-on” sensor. Fluorescence activates only upon Tau aggregation, with minimal background under basal conditions. In HEK293 cells expressing full-length Tau, aggregation induced by phosphorylation-promoting compounds produced strong fluorescence, enabling detection of early aggregation events.
Huttunen’s group employed a split GFP system with the 0N4R Tau isoform, although its functionality was not thoroughly investigated. Their study primarily focused on using the split GFP Tau system to examine the effects of conditioned media containing Tau on neighboring cells with respect to stress granules, rather than comprehensively validating its utility for detecting aggregation dynamics or high-throughput drug screening. This highlights the need for more robust and systematically validated systems, such as the one presented in this study, to effectively study Tau aggregation and identify therapeutic modulators.
Building on Huttunen’s approach, we are using a similar split GFP construct with 0N4R Tau but specifically optimizing the system for drug screening applications. Our goal is to identify compounds capable of preventing Tau-Tau interactions in a scalable and efficient manner. Importantly, while most studies on cell-based Tau aggregation rely on microscopy-based methods, which are time-consuming and challenging to quantify accurately, we have adapted our system for flow cytometry, offering rapid and reliable quantification and high throughput.
One notable exception to the reliance on microscopy is a recent study by Allsop and Mudher’s group, which used an ATCC cell line expressing K18 Tau to evaluate an aggregation inhibitor. However, their approach, like many others, uses adherent cells, which must be detached for flow cytometry analysis. Additionally, most publications report that Tau aggregation in cell-based systems requires the addition of seeds or other aggregation inducers. Even when truncated forms such as K18 mutants are used, aggregation without seeds is rarely observed. Allsop and Mudher’s work further noted that FRET signals in their system did not occur unless seeds were introduced.
In contrast, our system demonstrates that GFP fluorescence can be directly detected in suspension-adapted HEK293 cells co-transfected with split GFP constructs (GFP10 and GFP11) fused to 0N4R Tau. These suspension cells offer several advantages: they grow at higher densities, are less influenced by microtubule interactions, and produce greater amounts of protein. We hypothesize that this higher protein production enables the observation of Tau-Tau interactions in these cells, even without the need for adding aggregation inducers, a key difference compared to adherent HEK293 cells described in other studies. Furthermore, suspension cells are inherently better suited for flow cytometry-based applications, allowing for efficient, high-throughput screening of Tau aggregation inhibitors. This streamlined approach not only simplifies experimental workflows but also provides a robust and scalable platform for drug discovery.
