IntroductionAlzheimer-type neuropil threads (NTs) and neurofibrillary tangles (NFTs) are comprised of either 4 repeat (4R)-tau, 3 repeat (3R)-tau, or a mixture of both. In the hippocampus, the number of NFTs, and the proportion of 3R tau progressively increases. If this preferential accumulation of 3R tau also occurs in the brainstem, it may be fundamentally related to progression of Alzheimer pathology.MethodsMidbrain and pontine sections of brainstems from 23 cases (Braak-NFT stages I/II: 8, III/IV: 8, and V/VI: 7) were double immunofluorolabeled for 4R and 3R tau. High-resolution (0.645 μm/pixel), in-focus snapshots were tiled to cover entire brain sections using a virtual slide system. Each lesion was classified by size (NT < 200 μm2 < NFT) and staining profile (3R/4R). In addition, the localization and quantity of amyloid β (Aβ) deposits were examined in adjacent sections for comparison with tau.ResultsThe data sets obtained from approximately 286 gigabytes of image files consisted of 847,763 NTs and 7859 NFTs. The proportion of 3R tau-positive NTs and NFTs in the midbrain, and 3R tau-positive NTs in the pons gradually increased with advancing NFT stages, while the proportion of 3R tau-positive NFTs in the pons was already elevated at early stages. Aβ deposits were absent at NFT stages I/II, and when present at later stages, their regional distribution was different from that of tau. These observations suggest that a progressive increase in the proportion of 3R tau occurs independently of Aβ deposits.ConclusionsThis is the first quantitative analysis of NFTs and NTs in the human brainstem. We demonstrate that the proportion of 3R tau in the brainstem neurofibrillary changes increases with disease progression. Because this phenomenon is shared between the brainstem and the hippocampus, this increase may be fundamental to the pathogenesis of Alzheimer disease.Electronic supplementary materialThe online version of this article (10.1186/s40478-017-0501-1) contains supplementary material, which is available to authorized users.

No abstract

Neurofibrillary tangles (NFTs), are shared between progressive supranuclear palsy (PSP) and Alzheimer disease (AD). Histological distinction of PSP and AD is possible based on the distribution of NFTs. However, neuropathologists may encounter diagnostic difficulty with comorbidity of PSP and AD. In this study, we tried to circumvent this difficulty by analyzing five autopsied brains harboring both PSP and AD pathology. Tau-positive lesions were sorted based on their cell type (neuron versus glia), and tau isoforms: three-repeat (3R) versus four-repeat (4R) tau. 16 regions were selected to map these lesions throughout the brain. 4R-tau lesions were present in all areas examined. Among them, 3R-tau lesions were absent in some areas. These 4R selective (4R+/3R-) areas dictate prototypic distribution of PSP, not usually found in AD, such as pontine nucleus, red nucleus, inferior olivary nucleus, dentate nucleus, globus pallidus and putamen, each contained both glial and neuronal lesions. In contrast, additional 3R-tau lesions were found in hippocampal formation to neocortex, where 3R immunoreactivity (IR) was predominant over the 4R counterpart mainly in neurons as found in AD but not in PSP. Although tau lesions in central grey matter, substantia nigra and locus coeruleus are found in both AD and PSP, 4R-selectivity with glial component suggests PSP origin. Even if the presence of 3 R IR in these areas suggests AD pathology, it does not exclude the involvement of PSP-type lesion because distinction of 4R IR into PSP or AD is not yet possible. Further demixing may be possible if biochemical difference of 4R tau between PSP and AD is identified.