Development of a novel tau propagation mouse model endogenously expressing 3 and 4 repeat tau isoforms

Abstract

The phenomenon of ‘prion-like propagation’ in which aggregates of abnormal amyloid-fibrilized protein propagate between neurons and spread pathology, is attracting attention as a new mechanism in neurodegenerative diseases. There is a strong correlation between the accumulation or spread of abnormal tau aggregates and the clinical symptoms of tauopathies. Microtubule-associated protein tau (MAPT) contains a microtubule-binding domain that consists of three or four repeats (3R/4R) due to alternative mRNA splicing of transcripts for the MAPT gene. Although a number of models for tau propagation have been reported, most use 4R human tau transgenic mice or adult wild-type mice expressing only endogenous 4R tau and these models have not been able to reproduce the pathology of Alzheimer's disease in which 3R and 4R tau accumulate simultaneously, or that of Pick’s disease in which only 3R tau is aggregated. These deficiencies may reflect differences between human and rodent tau isoforms in the brain. To overcome this problem, we used genome editing techniques to generate mice that express an equal ratio of endogenous 3R and 4R tau, even after they become adults. We injected these mice with sarkosyl-insoluble fractions derived from the brains of human tauopathy patients such as those afflicted with Alzheimer’s disease (3R and 4R tauopathy), corticobasal degeneration (4R tauopathy) or Pick’s disease (3R tauopathy). At 8–9 months following intracerebral injection of mice, histopathological and biochemical analyses revealed that the abnormal accumulation of tau was seed-dependent, with 3R and 4R tau in Alzheimer’s disease-injected brains, 4R tau only in corticobasal degeneration-injected brains and 3R tau only in Pick disease-injected brains, all of which contained isoforms related to those found in the injected seeds. The injected abnormal tau was seeded, and accumulated at the site of injection and at neural connections, predominantly within the same site. The abnormal tau newly accumulated was found to be endogenous in these mice and to have crossed the species barrier. Of particular importance, Pick’s body-like inclusions were observed in Pick’s disease-injected mice, and accumulations characteristic of Pick’s disease were reproduced, suggesting that we have developed the first model that recapitulates the pathology of Pick’s disease. These models are not only useful for elucidating the mechanism of propagation of tau pathology involving both 3R and 4R isoforms, but can also reproduce the pathology of tauopathies, which should lead to the discovery of new therapeutic agents.