11 May 2005: In vitro generation of infectious scrapie prions

New work adds weight to the hypothesis that proteins alone are the infectious agent in transmissible spongiform encephalopathies (TSEs), including BSE, vCJD and scrapie.  

Stanley Prusiner’s controversial hypothesis, proposed in 1982, maintains that the infectious agent, or prion protein, is chemically identical to a healthy cellular glycoprotein, PrP^c, but in an abnormal form (PrP^res).   The insoluble, protease resistant PrP^res is self-propagating, causing the conformational changes in PrP^c  - a decrease in ahelix content and increase in bsheet.   Aggregation of abnormal protein in nerve cells creates brain lesions, and provides the source of infection.

Critics of the now widely accepted hypothesis point to genetic variation among TSEs - the distinctive pathologies of some 15 strains can be identified in a single model host, the laboratory mouse.   ’Such variation is readily explained if the protein is protecting a small information molecule, such as a nucleic acid,’ said Robert Somerville of the UK’s Institute of Animal Health in Edinburgh.   ’Although no such molecule has been found, it is unclear how conformational differences in PrP alone could encode the diversity of genetic information.’  

Generation of PrP^res in vitro,  isolated from disease material, would support a prion agent if the new protein itself was infective.   Claudio Soto and colleagues at the University of Texas Medical Branch, Galveston, US, explored this concept using protein misfolding cyclic amplification (PMCA).   Aggregates of PrP^res - formed when PrP^c is seeded with minute amounts of abnormal protein - are sonically fragmented and used to transform more PrP^c.  

Using brain homogenate from scrapie-infected hamsters solely to seed the first transformation, the technique was used to serially amplify and dilute PrP^res aggregates generated from the PrP^c of healthy brains.   Abnormal protein collected from a 10^-20  dilution of the original sample, was biochemically similar to the scrapie-brain PrP^res, and caused scrapie when injected into the brains of healthy hamsters.   ’Scrapie-brain homogenate diluted to 10^-20 is not infectious, and is mathematically devoid of PrP^res, so we consider PMCA and dilution removed all the original scrapie-brain PrP^res,’ said Soto.   ’Infectious PrP^res generated in vitro only from non-infectious material rules out the possibility of a virus-like agent with nucleic acid.’

Although Somerville concedes the results are promising, some reservations remain.  

The PMCA derived PrP^res was substantially less infectious than original scrapie material - experimental animals survived significantly longer than scrapie-brain infected controls.   The team’s various explanations for the anomaly, including a new strain generated during PMCA, are consistent with the prion hypothesis. However, Somerville asserts, ’the possibility remains that the experiment did not truly replicate the TSE infectious agent.   PrP^c was provided by brain homogenate so other reactions may have occurred during PMCA.   Moreover, the study operated near the limits of detection for PrP, in an examination of infectivity that is very difficult to destroy, so cross-contamination must be a concern.’

While the debate continues, PMCA might be applied elsewhere.   ’Its effectiveness in providing detectable PrP^res from minute samples could be used in a TSE blood test, much needed to improve surveillance for BSE and its human form, new-variant CJD,’ said Soto. 

Russ Clare