Determination of the frequency and distribution of vascular and parenchymal amyloid with polyclonal and N‐terminalspecific PrP antibodies in scrapie‐affected sheep and mice

Brains from 17 histopathologically confirmed cases of scrapie, five of which had congophilic vascular amyloid, were stained immunohistochemically for prion protein (PrP) using a polyclonal antibody. Two clinically suspect but pathologically unconfirmed cases of natural sheep scrapie and the brains of four mice infected with the 111A murine scrapie strain were also examined. Selected sections containing amyloid were stained with each of two peptide antibodies which recognise the N‐terminal amino acid residues which are lost following protease digestion of the disease‐specific isoform of PrP. The mice infected with the 111A murine scrapie strain had large numbers of hypermature plaques. All the amyloid plaques from both natural sheep scrapie brains and experimental murine brains were heavily immunostained by the polyclonal and both peptide antibodies. In addition, disease‐specific accumulations of PrP were detected in endothelial cells or in the intima of blood vessels of the cerebral cortex of sheep scrapie brains. The affected blood vessels were located in areas which otherwise lacked typical scrapie pathology. Vascular accumulations of PrP were also found in leptomeningeal and choroid plexus blood vessels. Vascular amyloid was found mainly in the neocortex. Vascular amyloid and disease‐specific parenchymal accumulations of PrP were found in two sheep which showed clinical signs of scrapie but lacked its typical vacuolar pathology. These results show that the mature amyloid of scrapie is composed of, or contains a substantial proportion of, whole length PrP protein. Thus truncation of PrP is not essential for the aggregation of PrP into amyloid. The vascular amyloid of natural sheep scrapie originates from the accumulation and release of PrP from endothelial cells presumably following systemic scrapie infection. The topography of vascular amyloid distribution in Great Britain differs from that reported in the Netherlands. As amyloid deposition in mice is largely controlled by the strain of the infecting agent it is possible that the strain of the agent may influence vascular amyloid deposition.