Tau Filaments Act Like Prions, Imposing Their Disease Signature on Mouse Brains

In a landmark study published in Nature, researchers from the Tokyo Metropolitan Institute of Medical Science and the MRC Laboratory of Molecular Biology in the United Kingdom have provided the most direct evidence yet that tau, a protein central to Alzheimer’s disease and a family of related dementias, behaves like a prion in the living brain. Led by Dr. Masato Hasegawa, Deputy Director of the Tokyo Metropolitan Institute of Medical Science, the team injected insoluble tau filaments extracted from the brains of patients with Alzheimer’s disease or corticobasal degeneration into the brains of ordinary, wild-type mice. What happened next was striking: the mouse brain’s own tau protein was recruited into filaments that copied, with near-perfect fidelity, the molecular architecture of the injected human seeds.

The concept of prion-like transmission has shadowed tau research for years. Prions, the proteinaceous infectious agents behind Creutzfeldt-Jakob disease and bovine spongiform encephalopathy, propagate by coaxing normally folded prion proteins into the abnormal, misfolded state. Distinct folding patterns of the abnormal protein behave as different prion strains, each carrying its own incubation period and clinical profile. Tau, a microtubule-binding protein implicated in more than twenty neurodegenerative diseases, forms amyloid-like filaments in patient brains and has long been suspected of spreading through a similar templating mechanism. Cryo-electron microscopy studies have already shown that each tauopathy is defined by a unique tau filament fold, and that the same fold recurs across different brain regions within a single patient, implying that specific structures amplify and propagate as disease advances.

What remained unproven was whether tau filaments of different structures genuinely cause distinct disease pathologies through prion-like mechanisms, and whether those structural signatures survive the journey from cell to cell. The new study, titled Prion-like transmission of human tau strains in the mouse brain and published on September 30, 2026, answers both questions with unusual clarity. The work was a collaboration involving Dr. Aki Shimozawa and Dr. Airi Tarutani in Tokyo alongside Dr. Sofia Lövestam, Dr. Michel Goedert, and Dr. Sjors H.W. Scheres in Cambridge, with Lövestam, Shimozawa, and Tarutani serving as co-first authors.

The experimental design was elegantly simple. Tau filaments were extracted from the brains of Alzheimer’s disease and corticobasal degeneration patients and injected into the striatum of wild-type mice aged six to eighteen weeks. No genetic engineering was involved; the animals carried only their own normal mouse tau. Within six to nine months after injection, the mice developed tau pathologies that closely resembled those seen in the patient brains from which the seeds had come. Immunohistochemical staining revealed that the pathology had spread well beyond the injection site, reaching the cerebral cortex, the corpus callosum, and other distant brain regions, mirroring the progressive spread of tau lesions in human disease.

A critical series of controls established that the accumulating pathology was not simply the injected material persisting in the brain. Time-course analysis showed that the human tau filaments degraded and disappeared within approximately one week of injection. Endogenous tau accumulation began to re-emerge one to three months later, and by six to nine months the disease-characteristic lesions had fully appeared. Immunoblotting of the sarkosyl-insoluble fractions, a biochemical preparation that isolates aggregated tau, showed that the insoluble tau in the mouse brains reacted with an antibody specific to mouse tau, called mTau, but not with HT7, an antibody specific to human tau. The pathology, in other words, was built entirely from the mice’s own protein, assembled under the direction of the transient human seeds.

The most consequential finding came from cryo-electron microscopy, a technique that freezes biological molecules in place and resolves their three-dimensional structures at atomic scale. Filaments harvested from the mouse brains adopted the exact same folded structures as the injected human tau filaments. In mice seeded with Alzheimer’s-derived tau, the majority of amplified filaments were helical structures composed of two twisted protofilaments, resolved at 3.6 angstroms and shown to be identical to the paired helical filaments characteristic of Alzheimer’s patient brains. In mice seeded with corticobasal degeneration-derived tau, two filament types emerged: roughly seventy percent were single protofilaments and thirty percent were doublets of two twisted protofilaments, with the single protofilaments resolved at 3.4 angstroms and proven identical to Type 1 filaments found in corticobasal degeneration patients.

Equally revealing was how the two seed types produced visibly different diseases in the same kind of host. Mice injected with Alzheimer’s-derived tau developed pathology concentrated in the cell bodies and processes of neurons. Mice injected with corticobasal degeneration-derived tau showed tau accumulation not only in neurons but also in glial cells, forming structures that resembled astrocytic plaques and coiled bodies, the hallmark lesions used to diagnose corticobasal degeneration in human neuropathology. Astrocytic plaques are tau deposits in the distal processes of astrocytes, while coiled bodies are fibrous, comma-shaped inclusions in oligodendrocytes. The structural identity of the seed therefore determined not just the shape of the filaments but which cell types would be affected and how the pathology would be distributed across the brain.

Immunoelectron microscopy of the insoluble fractions confirmed the picture, revealing abundant filament structures decorated by both mTau and AT8 antibodies, markers of mouse tau and pathological phosphorylation respectively. Further biochemical profiling with antibodies recognizing the C-terminal region of tau showed that the aggregates formed by Alzheimer’s-derived and corticobasal degeneration-derived seeds carried distinct biochemical signatures, reinforcing the conclusion that the injected structures had imposed their identities on the newly formed mouse tau. At the ultrastructural level, the Alzheimer’s-seeded mice produced filaments with the tight twisted configuration typical of that disease, while the corticobasal degeneration-seeded mice produced filaments with a longer twist.

The implications reach well beyond the laboratory. These findings provide concrete evidence that tau propagates through the brain the way prion strains do, using filament seeds as structural templates that convert normal protein into copies of themselves while preserving their characteristic folds through successive rounds of amplification and cell-to-cell transmission. Because the mouse model recapitulates disease-specific pathology without any genetic manipulation, it offers researchers a powerful and physiologically faithful platform for dissecting how different tau filament structures produce different diseases, and for testing interventions designed to block seeding, spread, or specific filament conformations.

For the millions of people affected by Alzheimer’s disease and related tauopathies, the study points toward a new class of therapeutic logic. If disease progression depends on tau strains templating their own propagation, then drugs that stabilize normal tau, block filament uptake, or disrupt the templating surface could halt the spread of pathology even after it has begun. The research was supported by the Tokyo Metropolitan Government, the Japan Agency for Medical Research and Development, the Japan Science and Technology Agency, and the Japan Society for the Promotion of Science KAKENHI program, and all animal experiments were approved by the Animal Care and Use Committee of the Tokyo Metropolitan Institute of Medical Science. Patient brain tissue was provided by collaborating institutions across Japan and the University of Manchester, a reminder that this advance in molecular neuroscience rests on the generosity of patients and families who donated tissue for research.

Subject of Research: Prion-like transmission of disease-specific tau filament structures in the mouse brain

Article Title: Demonstration of prion-like transmission of tau in the mouse brain—elucidating the mechanisms behind disease-specific pathogenesis—

Article References: Demonstration of prion-like transmission of tau in the mouse brain—elucidating the mechanisms behind disease-specific pathogenesis—. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: tau, prion-like transmission, Alzheimer's disease, corticobasal degeneration, cryo-electron microscopy, neurodegeneration, tauopathies, protein aggregation, mouse model, amyloid filaments, seeding, Nature