A beta and perlecan in rat brain: glial activation, gradual clearance and limited neurotoxicity

A beta 1-40 and perlecan (A beta + perlecan) were infused into rat hippocam pus for 1 week via osmotic pumps. At the end of the infusion a deposit of A beta immunoreactive material was found surrounding the infusion site. No n eurons could be identified within this A beta deposit. The neuron-free area resulting from A beta + perlecan was significantly larger than that found after infusions of A beta 40-1 and perlecan (reverse A beta + perlecan), pe rlecan alone or phosphate-buffered saline vehicle. Following infusion of A beta + perlecan, the glial cells segregated in a manner similar to that ass ociated with compacted amyloid plaques in Alzheimer's disease (AD). Activat ed microglia/macrophages were prevalent within the A beta deposit while the perimeter of the deposit was delimited by reactive astrocytes. Thioflavin S and Congo red staining indicated a beta-pleated sheet conformation of the A beta deposits, implying formation of fibrils. Intact, apparently healthy neurons were found immediately adjacent to the A beta + perlecan deposit. In contrast, reverse A beta peptide did not form congophilic deposits despi te the presence of perlecan. Apoptotic profiles visualized with bisbenzamid e or TUNEL staining of fragmented DNA were not seen at any of the infusion sites, yet were readily seen in hippocampal sections from animals treated w ith kainic acid. At 8 weeks, A beta immunoreactivity, Thioflavin S and Cong o red staining was reduced, indicating that A beta was being cleared. There also was no evidence of neuron loss by Nissl or TUNEL staining. The zone o f apparent necrosis did not expand between 1 and 8 weeks, and in some insta nces appeared to contract. The consistency of the A beta + perlecan infusio n method in producing reliable A beta amyloid deposits permits estimates of the rate at which fibrillar A beta amyloid can be removed from the brain, and may provide a useful model to study this process in vivo. However, the absence of clearly identifiable degenerating/dying neurons at the 1 or 8 we ek survival times suggests that either fibrillar A beta + perlecan slowly d isplaced the brain parenchyma during infusion, or neurons were killed very gradually during the process of clearing the A beta. (C) 2000 Elsevier Scie nce Ireland Ltd. All rights reserved.