17 Sep 2003

Prions, those self-replicating proteins that are responsible for scrapie, mad cow disease, and human neurologic diseases such as Creutzfeldt-Jakob and Kuru, have been notoriously difficult to pin down in the cell. Now, an international collaboration led by Peter Peters at the Netherlands Cancer Institute, Amsterdam, and Stanley Prusiner at the University of California, San Francisco, reports perhaps the most detailed localization study to date. Importantly, their data, which appeared in the August 6 Journal of Neuroscience, show that a pool of prions exists in the neuronal cytosol, supporting a recent hypothesis that minute amounts of misfolded, cytosolic prion may be sufficient to cripple neurons.

Normal, or cellular prion protein (PrPC), is an innocuous cell-surface molecule expressed throughout the brain. A Jekyll and Hyde-like conformational change in the protein turns it into a lethal infectious agent (PrPSc) that can self-propagate by catalyzing the conversion of other normal prions. How this protein kills neurons is unclear. However, Jiyan Ma and Susan Lindquist recently revealed that a third form of the prion may arise through occasional misfolding during synthesis. This group proposed that this misfolded protein is ejected from the endoplasmic reticulum into the cytosol, where the proteasome destroys it. But if the proteasome is overwhelmed, then this cytosolic prion (cytPrP) could wreak havoc on the cell, most likely by forcing it down the road to apoptosis and death (see ARF related news story, but also Roucou et al., 2003).

The latest findings from Peters et al. lend credence to this hypothesis. First author Alexander Mironov and colleagues used light and electron microscopic analysis of semi-thin and ultra-thin tissue sections from mouse brain to show that PrPC appears in expected locations, including the cell membrane, axons, pre- and postsynaptic boutons, and also in organelles involved in the synthesis of membrane-bound proteins, such as the endoplasmic reticulum and the Golgi complex. But Mironov also found that in a subset of CA1 hippocampal cells, the bulk of the prion occurred in the cytosol. Though the identity of these cells is unclear, they are most likely neurons, claim the authors, because they do not stain positive for glia or oligodendrocyte markers.

Many pieces of the prion puzzle remain. Though the cytPrP cells identified by Mironov have irregularly shaped nuclei and abundant dense cytoplasm, in all other respects they seem normal, with no signs of degeneration or activation of apoptosis. This contrasts the finding of Ma and colleagues, who found that even a minor amount of cytPrP can be lethal to cells. The answer to this conundrum may lie in specific cell types. Cerebellar cells appeared most affected when Ma et al. targeted PrP to the cytosol, but Mironov could not find cytPrP in that area of the brain. The authors conclude that “cytPrP is not toxic in some neurons but highly toxic when overexpressed in specific cell populations.” This suggestion is reminiscent of the pattern of destruction in other neurodegenerative diseases, where misfolded proteins are thought to play a major role. In Alzheimer’s and Parkinson’s, to name just two, neuronal destruction begins in specific cells in specific regions of the brain. Elucidating the nature of prion toxicity may, therefore, shed some light on other neurodegenerative diseases.—Tom Fagan