Tau seed amplification assay reveals relationship between seeding and pathological forms of tau in Alzheimer's disease brain.

Tau seed amplification assays (SAAs) directly measure the seeding activity of tau and would therefore be ideal biomarkers for clinical trials targeting seeding-competent tau in Alzheimer's disease (AD). However, the precise relationship between tau seeding measured by SAA and the levels of pathological forms of tau in the AD brain remains unknown. We developed a new tau SAA based on full-length 0N3R tau with sensitivity in the low fg/ml range and used it to characterize 103 brain samples from three independent cohorts. Tau seeding clearly discriminated between AD and control brain samples. Interestingly, seeding was absent in Progressive Supranuclear Palsy (PSP) putamen, suggesting that our tau SAA did not amplify 4R tau aggregates from PSP brain. The specificity of our tau SAA for AD brain was further supported by analysis of matched hippocampus and cerebellum samples. While seeding was detected in hippocampus from Braak stages I-II, no seeding was present in AD cerebellum that is devoid of tau inclusions. Analysis of 40 middle frontal gyrus samples encompassing all Braak stages showed that tau SAA seeding activity gradually increased with Braak stage. This relationship between seeding activity and the presence of tau inclusions in AD brain was further supported by robust correlations between tau SAA results and the levels of phosphorylated tau212/214, phosphorylated tau181, aggregated tau, and sarkosyl-insoluble tau. Strikingly, we detected tau seeding in the middle frontal gyrus already at Braak stage II-III, suggesting that tau SAA can detect tau pathology earlier than conventional immunohistochemical staining. In conclusion, our data suggest a quantitative relationship between tau seeding activity and pathological forms of tau in the human brain and provides an important basis for further development of tau SAA for accessible human samples.