Scrapie prion rod formation in vitro requires both detergent extraction and limited proteolysis

Scrapie prion infectivity can be enriched from hamster brain homogenates by using limited proteolysis and detergent extraction. Purified fractions contain both scrapie infectivity and the protein PrP 27-30, which is aggregated in the form of prion rods. During purification, PrP 27-30 is produced from a larger membrane protein, PrPSc, by limited proteolysis with proteinase K. Brain homogenates from scrapie-infected hamsters do not contain prion rods prior to exposure to detergents and proteases. To determine whether both detergent extraction and limited proteolysis are required for the formation of prion rods, microsomal membranes were prepared from infected brains in the presence of protease inhibitors. The isolated membranes were then detergent extracted as well as protease digested to evaluate the effects of these treatments on the formation of prion rods. Neither detergent (2% Sarkosyl) extraction nor limited proteinase K digestion of scrapie microsomes produced recognizable prion amyloid rods. Only after combining detergent extraction with limited proteolysis were numerous prion rods observed. Rod formation was influenced by the protease concentration, the specificity of the protease, and the duration of digestion. Rod formation also depended upon the detergent; some combinations of protease and detergent did not produce prion amyloid rods. Similar results were obtained with purified PrPSc fractions prepared by repeated detergent extractions in the presence of protease inhibitors. These fractions contained amorphous structures but not rods; however, prion rods were produced upon conversion of PrPSc to PrP 27-30 by limited proteolysis. We conclude that the formation of prion amyloid rods in vitro requires both detergent extraction and limited proteolysis. In vivo, amyloid filaments found in the brains of animals with scrapie resemble prion rods in their width and their labeling with prion protein (PrP) antisera; however, filaments are typically longer than rods. Whether limited proteolysis and some process equivalent to detergent extraction are required for amyloid filament formation in vivo remains to be established.