Based on experiments with yeast, different cell lines, a newly created strain of transgenic mice, and previous work (Ma & Lindquist, 2001), the authors propose the following mechanism for how sporadic prion diseases, as well as some familial forms, develop: On its way to the cell surface, wild-type prion protein (PrP) passes through the endoplasmic reticulum (ER) for folding and glycosylation. Some of it usually misfolds and is ejected back into the cytoplasm, where the proteasome promptly degrades it. If the proteasome becomes overwhelmed due to aging, stress, or increased misfolding of mutated PrP, misfolded PrP accumulates in the cytoplasm. When it reaches a certain concentration, it spontaneously converts to a conformation that not only induces conversion by further misfolded prion proteins, but also becomes so highly neurotoxic that it kills neurons-possibly by apoptosis-at levels where it is barely detectable.

In search of a unifying pathogenic mechanism for prion diseases, Ma et al. were puzzled by the fact that PrP^Sc, the self-propagating form of PrP that is widely considered a culprit, is not itself neurotoxic. (Similarly, demonstrating in-vivo neurotoxicity of Aβ in Alzheimer's disease is proving vexingly difficult and is considered a weakness in the amyloid hypothesis.)

To study the relationship between misfolding, the proteasome, PrP accumulation and toxicity, Ma et al first showed that certain cell lines expressing wild-type PrP died when treated with proteasome inhibitors. After proteasome inhibition, the cells contained increased levels of cytosolic PrP species that had their ER signal sequences clipped, indicating they had been in the ER. The accumulating PrP resembled PrP^Sc in detergent insolubility and protease resistance. By contrast, other accumulating cytosolic proteins, for example presenilin, did not kill the cells, nor did PrP when it was located in other areas of the cell.

The rate of conversion to this new conformation increased with the level of PrP expression. Further experiments showed that once some PrP had converted, it kept promoting the conversion of further PrP molecules even when proteasome activity resumed after a temporary block. This suggests that misfolding of certain proteins can have an ominously self-sustaining character.

When the researchers tried to establish stable lines of cells expressing a purely cytosolic form of PrP (cyPrP), they made a curious discovery. All attempts to make a cyPrP-expressing neuroblastoma line failed. These cells died, even though they readily formed lines expressing wild-type PrP. Ma et al. eventually were able to make such a line by using an inducible promoter. With cyPrP expression turned off, the neuroblastoma cells grew (although a tiny amount of constitutive cyPrP expression kept growth slow amidst many apoptotic cells), but turning cyPrP expression on quickly killed the cells. Strangely, fibroblast-derived cells appeared not to mind accumulating cytosolic PrP and readily formed stable lines, indicating that perhaps protein-protein interactions specific to a cell's metabolism make cyPrP toxic. In Alzheimer's, as in other neurodegenerative diseases, the specific pattern of neuronal death continues to puzzle scientists.

Is this relevant in vivo? To find out, Ma et al. made transgenic mice expressing misfolded cyPrP. The phenotype was startling. As analyzed by co-author Robert Wollmann, also at University of Chicago, the mice had severe neurodegeneration of a pattern that resembles closely that of transgenic mice producing mutant, disease-derived forms of PrP. Mice expressing cyPrP from one allele developed pathology at five to 12 months, those expressing it from both alleles did so at two months of age. The mice had massive neuronal loss and atrophy in the cerebellum, in a cell-autonomous pattern that suggested individual neurons were dying as a consequence of an internal change, not by infection from outside. The models also showed gliosis, as well as a pattern of pathology in brain but not heart or muscle that mimics human prion disease. Despite the pathology, however, the researchers were able to detect only trace amounts of cyPrP in these mice's brains, suggesting that misfolded cyPrP is extremely neurotoxic in quantities so minute that it may well have been present but eluded detection in studies of other models and human brain, the authors write.

The present papers do not address Alzheimer's disease. However, one of the questions this data raises is whether an as-yet undetected cytosolic Aβ might be toxic in AD. The mechanism of neurotoxicity in AD is still unclear. Related research on small oligomeric forms of Aβ also wrestles with the technical problem that these species cannot be measured in vivo. Echoing somewhat the phenotype of the mice describe here, one transgenic mouse expressing Aβ inside neurons had apoptotic neuronal death, neurodegeneration and reduced life span ( LaFerla et al. 1995.) The question of whether, and how, intracellular Aβ might contribute to pathogenesis is drawing increasing attention. At the Alzheimer congress last July in Stockholm, for example, Gunnar Gouras presented work on accumulation of Aβ inside human AD-vulnerable neurons. Finally, this work might stimulate new thought about how conflicting theories about "good" and "bad" roles of Aβ might both be right in that the peptide is not normally toxic but becomes so only after misfolding. -Gabrielle Strobel.

References
Ma J., Lindquist S. Conversion of PrP to a Self-Perpetuating PrP^Sc-like Conformation in the Cytosol. 2002. SciencExpress, 17 October. Abstract

Ma J., Lindquist S. Neurotoxicity and Neurodegeneration When PrP Accumulates in the Cytosol. Science. 2002 Nov 29;298(5599):1781-5. Abstract