Prion protein infection mechanism identified

The transmissible spongiform encephalopathies (TSEs), which include variant Creutzfeldt-Jakob Disease in humans, “Mad Cow” Disease in cattle and scrapie in sheep, are progressive and fatal neurodegenerative disorders characterized by the accumulation within nerve cells of an abnormally folded and insoluble form of the prion protein.

The infectious agent which causes these diseases is generally believed to be the prion protein itself. According to Stanley Prusiner’s prion hypothesis, abnormal prion molecules act as a “seed” upon their entry into a cell, causing the normal cellular form of the protein to adopt the abnormal configuration. To date, however, the mechanism by which the infectious prion particle is passed from cell to cell has remained unclear.

Now researchers from the University of San Diego School of Medicine and The Scripps Research Institute in La Jolla have identified the process by which the pathological form of prion protein infects cells. In a new study published in the open access journal PLoS One, they report that the prion protein shares some characteristics with an HIV protein called Tat, and enters cells by the same mechanism.

The normal cellular form of the prion protein is present in most or all neurons, and although its function is unknown, it can be inferred from the protein’s location within the cell. Prion molecules are found at the cell surface, where they are anchored to the nerve cell membrane by short fatty acid molecules called glycosylphosphatidylinositols (GPIs), which are attached to the prions following synthesis of the proteins. Prions are, therefore, most likely to play a role (or roles) in some aspect of the cell-to-cell signalling process.

Prion proteins are continuously recycled between the membrane and the endosomes, small membrane-bound compartments found within the cell. The mechanism by which prions are taken into the cell from the membrane is still unclear, although several mechanisms have been proposed. One would expect the GPI anchor to be involved, as it is in other cell surface proteins. In the case of prions, however, this appears to be mediated by a small segment of the protein itself, located near one end of the molecule: deletion of this segment, or a mutation within it, disrupts prion internalization, whereas fusion of the segment to other GPI-anchored proteins promotes it.

The authors of the new study had previously found that protein internalization directed by a short protein sequence called the peptide transduction domain (PTD) involves 3 steps. First, electrical charges on a PTD-containing protein interact with those of cell surface molecules called proteoglycans. The cell membrane then invaginates to create a pocket which is continuous with the outside of the cell and which contains the protein which is to be internalized. This pocket dissociates from the membrane and is pulled into the cell. Finally, the contents of the pocket are released into the cytoplasm. The HIV peptide Tat, which accelerates the production of viral particles and facilitates infection, enters cells in this way.   

After analysing the prion protein amino acid sequence and finding that it contains a putative PTD, the authors compared its uptake to that of the HIV peptide. They used genetic engineering to fuse each of them to a fluorsecent marker protein – the mouse prion protein was fused to red fluorescent protein, and Tat to green fluorescent protein. The labelled proteins were then incubated with N2a neuroblastoma cells. (This cell line is derived from the tumour of the same name, and is commonly used for cell biological studies.) When the live cells were observed under the microscope, both proteins were found to be rapidly localized to the same regions, suggesting that they are internalized in the same way. To further investigate the internalization mechanism, the authors changed slightly the incubation conditions. When the N2a cells and flurescently labelled proteins were incubated in the presence of negatively charged heparin, no red fluorescence was seen within the cells. Likewise, when cholesterol-rich membrane structures called lipid rafts were disrupted, the prion protein neither bound to the N2a cell membranes nor entered the cells.

The researchers then dissected the prion molecule to determine which amino acid residues are required for its internalization. Different regions of the protein were first fused to the enzyme Cre recombinase, and incubated again with N2a cells. If taken up by the cells, the fusion proteins would then be translocated into the nucleus, where the Cre enzyme would reshuffle the chromosomal DNA so that the gene encoding green fluorescent protein is activated; this would then lead to synthesis of that protein, causing the cell to emit fluorescence. These experiments showed that fusion proteins containing amino acid residues 23-29 or 23-90 of the prion protein entered cells as efficiently as Cre-Tat. Fusion proteins containing prion residues 100-110 entered cells far less efficiently, but proteins containing residues 30-90 did not enter at all. These results strongly suggest that residues 23-29 of the prion protein constitute a strong, positively charged transduction domain, and residues 100-110 a weaker one.

In a final set of experiments, the researchers investigated exactly how prion protein enters cells. N2a cells expressing a fluorescently tagged protein which marks small membrane invaginations called caveolae were treated with fluorescently labelled full-length prion protein. Under the microscope, there was very little overlap of the different coloured fluorescences, showing that prion protein did not enter the cells by a caveolar-dependent mechanism. A  Cre-prion fusion protein was then incubated with N2a cells expressing a non-functional form of dynamin, a protein involved in clathrin-mediated endocytosis, the process by which synaptic vesicles are pulled back into nerve cells after they have released their cargo of neurotransmitter molecules. The Cre-prion protein entered the cells regardless, as evidenced by the green fluorescence emitted following entry of the fusion protein into the nucleus. Thus, the possibility that prion protein enters cells by clathrin-mediated endocytosis wasalso eliminated.

How, then, does prion protein enter cells? It turns out to be by a mechanism called macropinocytosis. This involves the invagination and pinching off of a segment of cell membrane, and is used by cells to take up small amounts of liquids (macropinocytosis means “big cell drinking”). It is dependent on the activity of a ubiquitous motor protein called actin, which forms the filaments of the cytoskeleton and is involved in many different forms of cell motility in organisms as diverse as yeast and humans. The authors had already found that Tat fusion proteins enter cells in this way, and they now confirm that this is the mode of entry for the prion protein too. They did so by blocking micropinocytosis and actin microfilament elongation separately – both treatments prevented the entry of fluorescently labelled prion protein into cells. Significantly, micropinocytosis was found to be necessary for the conversion of the normal cellular prion protein to the pathogenic abnormally folded form.


Wadia, J. S. et al (2008). Pathologic prion protein infects cells by lipid-raft dependent macropinocytosis PLoS ONE 3 (10) DOI: 10.1371/journal.pone.0003314.


  1. #1 Jen
    October 8, 2008

    So (if I understand this), either the actin microfilaments are elongated to allow prions entry, or micropinocytosis is inhibited to allow the prions entry, and then micropinocytosis must be present or reactivated to allow the prions to fold pathogenically? If the inhibition-entry guess is right, then the switch from inhibiting micropinocytosis to reactivating it is intriguing. If there isn’t an easy mechanism, then Occam’s razor would suggest elongating the actin filaments was right. How about this as a switch mechanism: like the prion finds a cell about to take a drink, rubs its protein on the proteoglycans, which makes the surface tension on the partial invagination seem such that it’s holding a mouthful of water, et voila, they’re in like Flynn? And micropinocytosis will resume because the “thirst” will still be present?

  2. #2 Mo
    October 8, 2008

    Perhaps parts of that last paragraph are unclear. Inhibiting micropinocytosis or actin filament polymerization prevented the prion protein from entering the cells.

  3. #3 Jen
    October 8, 2008

    Oops. I missed the “inhibiting” with the actin filament polymerization. I’m no doubt reading the following wrongly as well. In your blog entry you say
    The authors had already found that Tat fusion
    proteins *enter* cells in this way, and confirmed
    that prion protein *also does so by inhibiting either
    micropinocytosis* itself, or the elongation of actin

    I read that as inhibiting grants entry. I try to read it as prevents entry but can’t. I wish I knew what you meant there.

  4. #4 Mo
    October 8, 2008

    It made sense after the next paragraph:

    …both treatments prevented the entry of fluorescently labelled prion protein into cells

    I’ve edited it slightly anyway. Sorry about the confusion.

  5. #5 Jen
    October 9, 2008

    Aha! I was mixing up method and conclusion. Thanks for the explanation. This stuff is fascinating. Between molecular chaperones and now micropinocytosis, and TAT, and prion protein, I am aghast at how inventive and creative life is. To me, reading your blog is sometimes like getting glimpses of warring factions fighting on the side of an Everest on another world. Thanks!

  6. #6 eddie
    October 12, 2008

    An awesome experiment. You have cancer cells, invaded by a mixture of hiv and mad cow disease with fluorescent markers, to try and see how the infection gets into the cells.
    What’s the chance that andy schlaffly will demand samples of your cultures?

  7. #7 eddie
    October 12, 2008

    I’m most interested in the idea that the prions use similar method to the virus to invade cells, with the outcome that they make more copies of themselves.
    Is this a glimpse of non(or pre)-dna life?

  8. #8 Regis
    October 13, 2008

    Excuse the neophyte question. But somebody have to ask the question. Is this study validate that the prion acts “just” like an infectious agent, which would go in favor of the unknown virus?