Beta-protein/A4 deposits are not associated with hyperphosphorylated tau in somatostatin neurons in the hypothalamus of Alzheimer's disease patients.

With respect to the pathogenesis of Alzheimer's disease (AD), it has been hypothesized that amorphous plaques containing beta-protein/A4 (Abeta) would locally induce cytoskeletal changes, and that neurons affected by neurofibrillary tangles (NFTs) lose their neuropeptide concentration and eventually die. To test this presumed cascade of events, the hypothalami of 14 non-demented subjects (Braak 0-III) and 28 AD patients (Braak IV-VI) aged 40-98 years were selected. The subject of our study was the nucleus tuberalis lateralis (NTL), which harbors a subpopulation of somatostatinergic neurons with extensive intrinsic interconnectivity. We used Gallyas silver staining, Congo staining, single- and double-staining with monoclonal antibody AT8 and polyclonal antibody anti-Abeta, and double-immunolabeling with AT8 and anti-somatostatin(1-12) with the following results: (1) Significant amounts of silver-staining NFTs were present in only three AD patients. (2) High densities of AT8-stained cytoskeletal changes were mainly found in aged, demented patients. (3) In contrast, large amounts of Abeta deposits were mainly observed in young and middle-aged (40-59 years) AD patients, and were very low or absent mainly in the older non-demented subjects and in AD patients. (4) Reduced anti-somatostatin staining was observed in the NTL of most AD patients, but anti-somatostatin/AT8 double-stained neurons were found virtually exclusively in aged AD patients. Thus, the occurrence of Abeta deposits and hyperphosphorylated tau formation in somatostatin cells are basically independent events, while decreased somatostatin staining only partly goes together with cytoskeletal changes in somatostatin cells in the NTL of AD patients. These observations cannot be explained by the amyloid cascade hypothesis.