Polyubiquitin Repositioning Reshapes Pathological Tau Fibril Structures in Dementia

The pathological proteins that accumulate in the aging human brain are not amorphous debris. They are highly ordered molecular assemblies, and their precise atomic architecture can determine which neurodegenerative disease a patient develops, how fast it progresses, and how it spreads through neural circuits. A new study published in Nature Structural & Molecular Biology by Watanabe and colleagues adds a striking twist to this picture, showing that a modification located outside the ordered core of tau filaments can fundamentally remodel the filament’s internal structure. The finding suggests that the ultrastructure of disease-associated tau aggregates is not fixed by the tau sequence alone, but can be actively reshaped by post-translational modifications such as polyubiquitination.

Tau is a microtubule-associated protein that normally stabilizes the internal scaffolding of neurons. In Alzheimer’s disease and a family of related disorders collectively termed tauopathies, tau detaches from microtubules, misfolds, and polymerizes into paired helical and straight filaments that populate neurofibrillary tangles. For decades, these filaments were viewed through the lens of light and conventional electron microscopy, which revealed their overall morphology but not the atomic contacts that hold them together. The advent of cryo-electron microscopy changed that. By freezing thousands of filament samples in vitreous ice and computationally averaging images of countless molecular copies, researchers have resolved the ordered cores of tau filaments from Alzheimer’s disease, chronic traumatic encephalopathy, corticobasal degeneration, Pick’s disease, and several rarer conditions, each with a distinct fold.

These structures established a powerful principle: the fold of the filament core acts as a molecular signature of disease. Alzheimer’s disease tau filaments adopt one characteristic cross-beta architecture, while the filaments of other tauopathies adopt different folds, even though the tau protein sequence is the same. The core, however, is only part of the story. Large regions of the tau molecule remain disordered and invisible in cryo-EM maps, and the filaments carry a dense cargo of post-translational modifications, including phosphorylation, acetylation, truncation, and ubiquitination, many of which sit outside the resolved core. Whether these peripheral modifications merely decorate the filament or actively participate in determining its structure has been an open question.

The new work addresses this question directly by comparing tau filaments from Alzheimer’s disease with those from vacuolar tauopathy, a rare and recently characterized neurodegenerative condition marked by vacuolar changes in the brain and abundant tau inclusions. Using cryo-electron microscopy, the researchers determined the structures of tau filaments extracted from the brains of affected individuals, and then examined how these filaments behave in biological systems. A central result of the study is that tau fibrils from Alzheimer’s disease and from vacuolar tauopathy exhibit distinct seeding patterns in vivo. In other words, when these two types of filaments act as templates for the recruitment of soluble tau, they do not induce the same downstream aggregation behavior, indicating that their structural differences carry functional consequences for the propagation of pathology.

The most surprising structural insight concerns polyubiquitin. Tau filaments in many tauopathies are heavily ubiquitinated, a modification generally interpreted as a cellular tagging system that marks aggregates for degradation by the proteasome or for clearance by autophagy. In the filaments of vacuolar tauopathy, the researchers found that polyubiquitin chains are positioned differently relative to the filament core than in Alzheimer’s disease filaments. Crucially, this repositioning of polyubiquitin is associated with a shift in the protofilament interface, the set of contacts through which the two protofilaments that make up the fibril grip one another. A change at this interface means a change in the overall fold of the filament, altering which tau residues face the solvent and which are buried in the interior of the assembly.

This observation carries significant mechanistic weight. It demonstrates that a covalent modification attached outside the ordered core can exert enough structural influence to remodel the core itself. The polyubiquitin chains, though not part of the cross-beta spine, appear to constrain or redirect how protofilaments associate, effectively selecting an alternative filament fold. From a structural biology standpoint, this expands the concept of the tau filament from a self-determined protein polymer to a composite assembly whose architecture depends on both the tau sequence and the modifications it accrues in the diseased brain. The ordered core and its peripheral cargo are not independent; they are structurally coupled.

The in vivo seeding experiments reinforce the biological relevance of these structural differences. Seeding refers to the ability of an aggregate to template the misfolding of soluble tau, and it underlies the stereotyped spread of tau pathology through the brain along neural connections. If Alzheimer’s disease and vacuolar tauopathy filaments seed differently in living tissue, then the structural variants identified by cryo-EM are not laboratory curiosities but determinants of how disease propagates. Distinct folds may preferentially recruit distinct tau conformations, replicate with different efficiencies, or encounter different barriers to cellular uptake and intercellular transmission. This provides a structural framework for understanding why different tauopathies follow different clinical and pathological courses despite sharing the same aggregating protein.

The findings also have implications for therapeutic development. Several experimental treatments for Alzheimer’s disease and related disorders aim to clear tau aggregates or block their seeding, including immunotherapies with anti-tau antibodies and small molecules designed to stabilize non-pathogenic tau conformations. If the pathological filament structure varies between diseases, and even between molecular subpopulations within a single disease depending on ubiquitination state, then therapies will need to account for this structural heterogeneity. An antibody or inhibitor optimized against the Alzheimer’s disease filament fold may bind poorly to the vacuolar tauopathy fold, and vice versa. Conversely, the exposed polyubiquitin chains themselves could represent disease-specific epitopes or drug targets, offering a way to selectively recognize or destabilize particular filament variants.

More broadly, the study contributes to an evolving view of neurodegenerative disease in which the structural strain of an aggregate, not merely the identity of the aggregating protein, defines the disease entity. This concept, sometimes described as the prion-like strain hypothesis, has gained support from the growing catalogue of distinct filament folds resolved by cryo-EM across amyloid-beta, alpha-synuclein, and TDP-43 aggregates as well as tau. The demonstration that polyubiquitin repositioning can shift the protofilament interface adds a new layer to this framework: the strain landscape is not only diverse but malleable, shaped by the cellular environment and the modification state of the protein. Changes in the ubiquitin-proteasome system that accompany aging or disease could therefore, in principle, nudge tau filaments between alternative structural states.

Many questions remain. The precise atomic details of how polyubiquitin contacts the filament surface and transmits its influence to the protofilament interface will require further high-resolution analysis, and it is not yet clear whether similar modification-driven remodeling occurs in other tauopathies or in filaments carrying other modifications such as phosphorylation. It also remains to be determined whether repositioned polyubiquitin alters filament stability, clearance rates, or interactions with cellular quality-control machinery. Nevertheless, the central conclusion stands on firm ground: modifications outside the ordered core can remodel tau fibril ultrastructure, and the resulting structural variants seed differently in vivo. For a field that has long treated pathological filaments as static end products of neurodegeneration, the message is that these assemblies are dynamic, environmentally responsive structures whose architecture is written jointly by the protein sequence and the cell’s modification machinery. Understanding that interplay may prove essential for diagnosing tauopathies accurately and for designing interventions that target the right molecular shape in the right disease.

Subject of Research: Structural remodeling of pathological tau filaments by polyubiquitin repositioning in Alzheimer's disease and vacuolar tauopathy

Article Title: Repositioning of polyubiquitin alters the pathologic tau filament structure

Article References: Watanabe, R., Creekmore, B. C., Darwich, N. F., Smith, C. L., Xu, H., Baltazar, A., Salphati, S., Changolkar, L., Hoxha, K., Zhang, B., O’Rourke, C. M., Burslem, G. M., Lee, V. M.-Y., Chang, Y.-W., & Lee, E. B. (2026). Repositioning of polyubiquitin alters the pathologic tau filament structure. Nature Structural & Molecular Biology. https://doi.org/10.1038/s41594-026-01879-4

Image Credits: AI Generated

DOI: 10.1038/s41594-026-01879-4

Keywords: tau filaments, polyubiquitin, cryo-electron microscopy, Alzheimer's disease, vacuolar tauopathy, protofilament interface, neurodegeneration, seeding, post-translational modification, protein aggregation, amyloid strains, structural biology