Scientists’ view of amyloid’s role in Alzheimer’s disease has evolved from the view that large amyloid fibrils poison the brain to one primarily blaming soluble, oligomeric forms of Aβ. A similar evolution of thought may be on the horizon for prion diseases. In the February 24 Nature, researchers led by John Collinge at University College London, U.K., show that prion infection and propagation can be separated from toxicity. Using mice that expressed different levels of native prion protein (PrPc), the researchers found that prion infections rose rapidly and plateaued at the same level in all mice. Clinical disease began shortly thereafter in mice with a lot of prion protein, but much later in mice with little protein. The results suggest that prion disease involves two separate phases—an infectious phase and a toxic phase—with toxic prion forms emerging only after the infection has saturated.

In an accompanying Nature commentary, Reed Wickner at the National Institutes of Health, Bethesda, Maryland, wrote, “The authors infer that the toxic, killer PrP species must be distinct from the infectious species. This is unexpected and raises issues central to our understanding of prion diseases.” Lary Walker at Emory University in Atlanta, Georgia, agreed that the paper is conceptually interesting, adding, “We’re going to be thinking about the implications for quite awhile.”

In other prion news, researchers led by Lawrence Goldstein at the University of California in San Diego provide insight into prion transport in the February 18 Cell, identifying kinesin-1 and cytoplasmic dynein as the major motor proteins that tow prion-containing vesicles along axons.

Prion diseases are thought to develop when misfolded proteins infect an animal and catalyze the conversion of the native prion protein into the misfolded form. This mechanism has also captured the interest of AD researchers, since growing evidence suggests that aggregating forms of Aβ can convert harmless versions into toxic entities, thus spreading the disease through the brain (see e.g., ARF related news story on Eisele et al., 2009). Prion diseases involve long incubation periods, as long as 50 years in humans, followed by rapid clinical decline and death. In mice, researchers had noted that mice with more of the normal prion protein had shorter incubation periods than mice with less, but they did not know how the prion infection was progressing during this dormant period.

To answer this question, first author Malin Sandberg used a cell-based assay that measures the concentration of infectious prion particles accurately and rapidly (see Klöhn et al., 2003). Sandberg and colleagues compared brain homogenates from four different mouse strains: animals that lacked prion protein, hemizygous mice that expressed half the normal amount of prion protein, wild-type mice, and transgenic mice with eight times the normal expression level. All animals were inoculated with infectious Rocky Mountain Laboratory prions. At regular intervals, the researchers sacrificed animals and checked the infection level in their brains using the cell-based assay.

Mice without prion expression quickly cleared the infection, in accordance with prior research in the field (see, e.g., Büeler et al., 1993). Unexpectedly, the authors found that prion infection rose rapidly in all other mice, reaching a common plateau level that was independent of the amount of native prion protein in the brains. The infection phase took a little longer in animals with half the normal levels of prion expression, but was virtually the same length in wild-type mice and overexpressors. This suggests that something other than prion concentration limits the reaction, the authors write. It is not clear what sets the maximum level of infection; the authors speculate that there might be an essential cofactor that has a fixed concentration in the brain, or a limited number of sites for misfolded prions to occupy.

After the infections reached the plateau, the time until mice got sick varied inversely with the amount of prion protein the mice expressed. This fits with a model in which the infectious PrPsc particles act as a catalytic surface for production of neurotoxic PrP particles, the authors suggest. The toxic PrP species might be oligomeric, they add. In this model, mice die when poisonous prion particles reach a set toxic threshold that is the same for all mice. The more native prion lying around to serve as raw material, the more quickly the threshold is reached.

In his commentary, Wickner puts forth other possibilities. He points out that Sandberg and colleagues were unable to measure the concentration of toxic prion particles. Therefore, it could be that a higher level of toxic particles accumulates during the long incubation period in hemizygous mice, but these mice have more resistance to the deadly prions. To answer this question will require the development of a cell-based toxicity assay, Wickner suggests.

He also notes that only cells that express PrP on their surface are susceptible to the disease (see, e.g., Brandner et al., 1996). “In this sense, therefore, PrP is a receptor for killing cells,” Wickner wrote. “This would explain a direct relationship between toxicity and PrP expression level.” PrP particles may be assembling into amyloid forms on the cell surface, then getting internalized and clogging up cellular compartments, Wickner speculates.

That fits with previous work from Collinge’s group showing that depleting normal prion protein in mice allows them to recover from a prion infection (see ARF related news story on Mallucci et al., 2007; ARF related news story on Mallucci et al., 2003). These data highlight that the disease cannot take hold without the presence of normal prion protein.

Walker points out that the existence of a two-phase disease process suggests a therapeutic strategy. If physicians could catch the disease during its second, incubation phase, and administer a drug that lowers production of toxic prion forms, it might delay the onset of the disease. Walker said the next step will be to identify the infectious and toxic prion forms and discover how they differ from one another. If other diseases that involve protein aggregation, such as Alzheimer’s disease, also have this two-stage characteristic, the same approach might work for them, Walker speculated, adding that this is a big “if.” For AD, scientists must also gain a better understanding of what different Aβ species do, he said.

Meanwhile, new work from the UC San Diego group opens a window into the intracellular transport of normal prions. Goldstein and colleagues show that kinesin-1 and dynein are the major motor proteins responsible for carting prions around. Kinesin-1 powers anterograde transport toward axon ends, while dynein walks retrogradely, but both motors were always found on prion vesicles regardless of which way the vesicle was moving. The activities of both motor proteins were tightly coupled. This led the authors to suggest that the direction of movement must be controlled by the regulation of motor protein activity, rather than the attachment of specific motor proteins, Goldstein and colleagues found.—Madolyn Bowman Rogers


  1. With their recent paper on the transport of prion protein-containing vesicles, Encalada and her collaborators from Larry Goldstein’s laboratory made big steps in the right direction toward explaining how directionality of transport along axonal microtubules might be regulated in mammalian neurons. For us, "the right direction" means the right choice of the system of study and the tools employed. Few will disagree with the idea that transport is best studied by directly visualizing—in real time—the vesicles in motion. Also, few will disagree with the idea that this visualization should be done with minimal disruption of the intracellular environment. Practically, these two conditions are difficult to reconcile. First, visualization of a protein in motion implies that it is expressed as fusion protein, tagged with a fluorescent tracer, usually one of the many green fluorescent protein variants available today. In most cases, both the exogenous expression and the tag create problems. Ideally, one should look at the transport of an endogenous protein by the endogenous transport machinery, under native conditions. Transfection usually leads to expression of the tagged protein at above-normal levels, saturating the transport machinery, and the tag may interfere with the recruitment of the motors to the transport vesicle. Encalada et al. circumvent both problems by cleverly choosing the cellular prion protein (PrPC) as the prototype cargo they want to visualize. PrPC is a GPI-anchored glycoprotein, transported along the secretory pathway inside the carrier vesicles; this means that the tag—yellow fluorescent protein (YFP)—is also inside the lumen of the vesicle. In this way, the cytoplasmic face of the vesicle, which interacts with the molecular motor complexes, is not altered in any major way. The second clever thing that Encalada et al. do is to study the recruitment of the endogenous motors to the transport vesicles. Yes, they analyze motility of YFP-tagged PrPC with kymographs, but they use "vesicle mapping"—essentially immunocytochemistry of fixed specimens—to assess (literally, quantify) the number of motors associated with each vesicle, after prior recording of vesicle motility. The lesson from this study is that one should never underestimate the power of immunocytochemistry as a quantitative tool, and to work with endogenous, rather than expressed, proteins.

    The focus of these experiments was not to study transport of PrPC, but to address the question of how directionality of cargo vesicle transport along microtubules is regulated. In axons, this refers to the regulation of anterograde—plus-end-directed—versus retrograde—minus-end-directed—motility. Nerve ligation experiments undoubtedly indicate that the vesicles containing the PrPC move overall anterogradely, toward axon terminals. Yet, en route to the synapse, over shorter periods of time, many of them also move retrogradely, switching between anterograde and retrograde movement. Interestingly, Encalada et al. find that this switch in the direction of movement is enabled not by alternative release and recruitment of motors of opposed directionality, but by modulating, in a coordinated fashion, the activity of kinesin-1, the plus-end motor, and of cytoplasmic dynein, the minus-end motor, while both remain attached to the vesicle. Thus, the activities of kinesin-1 and cytoplasmic dynein seem tightly coupled, and the inactivation of one motor or another—done in this study with RNAi or with gene-targeted deletions of motor subunits—changes the direction of movement in unpredicted fashions. For example, the disruption of kinesin-1 not only decreases anterograde motility, but also the retrograde motility of PrPC-containing vesicles. This unexpected result contradicts the proposed tug-of-war mechanism, where the direction of vesicle movement is determined by the "stronger" team of opposing vesicle motors. It is not yet clear how the simultaneous regulation of the activities of the two types of motors is achieved. Encalada et al. envision some sort of physical interaction between kinesin-1 and cytoplasmic dynein. In our view, it is likely that scaffolding proteins capable of simultaneously binding kinesin-1 and cytoplasmic dynein could play a role. One such adaptor could be the Jun kinase (JNK)-interacting protein-1 (JIP-1), which enables the recruitment of kinesin-1 to vesicles carrying another famous cargo protein, the amyloid-β precursor protein (APP) (1,2). Recently, Erika Holzbaur’s group at the University of Pennsylvania showed that JIP-1 also binds a subunit of dynactin (3), an accessory protein complex of cytoplasmic dynein. Also, the C. elegans homologue of another JNK-interacting protein, JIP-3, was shown to bind both kinesin-1 and cytoplasmic dynein, enabling their co-transport into axons (4). Although vesicle association of this tripartite complex was not reported, we previously showed that JIP-3 is associated with cargo vesicles in mammalian neurons (5).

    On the same note, DISC1, a protein shown to bind at the same time kinesin-1 and cytoplasmic dynein (6), could be an example of an adaptor protein that recruits antagonistic molecular motors to the same vesicle. Intriguingly, DISC1 seems to be implicated in the transport of the APP-derived, amyloid-β (Aβ) peptide (7,8), a cargo that—like the prion protein studied by Goldstein’s group—is invisible to cytoplasmic proteins, being encapsulated within the carrier vesicle. Aβ is possibly tethered to the inner leaflet of the vesicle membrane via hydrophobic interactions. It may turn out that, rather than being passive scaffolds that recruit the motors to transport vesicles, adapter proteins such as JIP-1, JIP-3, or DISC1 could also regulate the coordinated activation and inactivation of the motors. The detailed molecular mechanisms that regulate the coordination and cooperation between motors of opposite direction of movement during vesicle transport remain to be elucidated. Certainly, such regulation is essential to allow the delivery of specific cargoes to their destination precisely when they are needed.

    Whether the mechanism of regulation of directionality of transport described by Encalada et al. also applies to vesicles transporting other cargoes remains to be seen. Override (9) and tug-of-war, mechanisms (10) could regulate transport of other cargo vesicles, or may operate in other cell types. As for the relevance for prion disease, and for neurodegenerative diseases in general, the contribution of the study of Encalada et al. is enormous. While it remains to be determined whether a deficient axonal transport is a cause, a contributing factor, or a consequence of the neuronal pathology in these diseases, there is no question that axonal transport is indeed abnormal in many neurodegenerative diseases (11). What is deregulated (e.g., motor recruitment to cargo, motor activation, their coordinated function, or the microtubule tracks), and how could the abnormality be corrected will only be known from studies like the one by Encalada and collaborators.

    References: See also Fu, M. and E.L. Holzbaur, JNK-induced changes in axonal transport are mediated by the bidirectional co-regulator JIP1. 50th Annual Meeting of the American Society for Cell Biology, Philadelphia, December 11-15, 2010; Muresan, V., B.T. Lamb, and Z. Muresan, DISC1 is required for the formation of intracellular Abeta oligomers, suggesting a link between schizophrenia and Alzheimer’s disease. Annual Meeting of the Society for Neuroscience, San Diego, November 13-17, 2010; and Muresan, V. and Z. Muresan, Disrupted-in-Schizophrenia 1 (DISC1) Facilitates the Intracellular Generation and Aggregation of Amyloid-beta (Abeta): A Link between Schizophrenia and Alzheimer’s Disease. 50th Annual Meeting of the American Society for Cell Biology, Philadelphia, December 11-15, 2010.


    . Amyloid beta protein precursor (AbetaPP), but not AbetaPP-like protein 2, is bridged to the kinesin light chain by the scaffold protein JNK-interacting protein 1. J Biol Chem. 2003 Oct 3;278(40):38601-6. PubMed.

    . Coordinated transport of phosphorylated amyloid-beta precursor protein and c-Jun NH2-terminal kinase-interacting protein-1. J Cell Biol. 2005 Nov 21;171(4):615-25. PubMed.

    . The Caenorhabditis elegans JIP3 protein UNC-16 functions as an adaptor to link kinesin-1 with cytoplasmic dynein. J Neurosci. 2011 Feb 9;31(6):2216-24. PubMed.

    . c-Jun NH2-terminal kinase-interacting protein-3 facilitates phosphorylation and controls localization of amyloid-beta precursor protein. J Neurosci. 2005 Apr 13;25(15):3741-51. PubMed.

    . DISC1 regulates the transport of the NUDEL/LIS1/14-3-3epsilon complex through kinesin-1. J Neurosci. 2007 Jan 3;27(1):15-26. PubMed.

    . Plus-end motors override minus-end motors during transport of squid axon vesicles on microtubules. J Cell Biol. 1996 Oct;135(2):383-97. PubMed.

    . Motor coordination via a tug-of-war mechanism drives bidirectional vesicle transport. Curr Biol. 2010 Apr 27;20(8):697-702. PubMed.

    . Is abnormal axonal transport a cause, a contributing factor or a consequence of the neuronal pathology in Alzheimer's disease?. Future Neurol. 2009 Nov 1;4(6):761-773. PubMed.

    View all comments by Virgil Muresan
  2. Malin Sandberg et al. show that the buildup of infectious prion particles (PrPsc) is distinct from the generation of toxicity, which then presumably is conferred by another PrP species (which we can call toxic PrP). There are clearly possibilities for mechanistic analogies in AD and other neurodegenerative diseases.

    The authors measure PrPsc levels in mice with different PrP expression after inoculating these with PrPsc. Mice that do not express PrP do not produce PrPsc, as expected. However, three other mice in which relative PrP expression ratios of 0.5 to 1 to 8, that is, over a range of a factor of 16, all build up PrPsc with similar rates and up to the same plateau concentration.

    At this point, some other process kicks in, which logically ought to be production of toxic PrP. This time, mice living with plateau concentrations of PrPsc become sick and die following plateau incubation times that are inversely proportional to the rates of PrP expression! In the language of physical chemistry, this is first-order kinetics in which the rate of toxic PrP generation is directly proportional to PrP concentration (given that all mice die at the same levels of toxic PrP). Hence, one can arrive at the following conclusions: 1) the production of infectious PrPsc and toxic PrP are separate processes; 2) there is some mechanism that limits infectious PrPsc to a maximum concentration; and possibly 3) toxic PrP starts to be generated when PrPsc reaches the maximum concentration. The evidence for number 3 is indirect since the levels of toxic PrP cannot be measured directly.

    The study leaves a few open questions that do not really cloud the story and that eventually may be answered. First, as pointed out in a comment by Reed Wickner in the same issue of Nature, the authors assume that the same amount of toxic PrP kills all types of mice. This might not be the case. Second, it is not completely clear exactly when toxic PrP starts to appear. Third, it would be nice to eventually find out how toxic effects are generated or mediated.

    Still, the possibility that similar or identical mechanisms operate in AD and other neurodegenerative diseases is intriguing. For instance, what about the long "plateau time" during which amyloid-β plaques accumulate before onset of AD, or even MCI symptoms? Are the plaques analogous to PrPsc and propagate from some kind of infectious seed? Is the delay between plaque appearance and onset of symptoms a period during which toxic Aβ species accumulate at a rate which is proportional to Aβ production? Is there, then, a critical concentration of toxic Aβ that triggers disease? Why, then, is this concentration so critical? Is it perhaps related to the initiation of tau pathology? How do we address these questions?

    These and related issues have been food for much thought already, and the new results on PrP will fuel the discussion.

    View all comments by Torleif Hard

Make a Comment

To make a comment you must login or register.


News Citations

  1. Aβ the Bad Apple? Seeding and Propagating Amyloidosis
  2. Early Intervention Reverses Prion Disease
  3. A Potential Prion Therapy Focuses Attention on Protein Conversion

Paper Citations

  1. . Induction of cerebral beta-amyloidosis: intracerebral versus systemic Abeta inoculation. Proc Natl Acad Sci U S A. 2009 Aug 4;106(31):12926-31. PubMed.
  2. . A quantitative, highly sensitive cell-based infectivity assay for mouse scrapie prions. Proc Natl Acad Sci U S A. 2003 Sep 30;100(20):11666-71. PubMed.
  3. . Mice devoid of PrP are resistant to scrapie. Cell. 1993 Jul 2;73(7):1339-47. PubMed.
  4. . Normal host prion protein necessary for scrapie-induced neurotoxicity. Nature. 1996 Jan 25;379(6563):339-43. PubMed.
  5. . Targeting cellular prion protein reverses early cognitive deficits and neurophysiological dysfunction in prion-infected mice. Neuron. 2007 Feb 1;53(3):325-35. PubMed.
  6. . Depleting neuronal PrP in prion infection prevents disease and reverses spongiosis. Science. 2003 Oct 31;302(5646):871-4. PubMed.

Further Reading

Primary Papers

  1. . Prion propagation and toxicity in vivo occur in two distinct mechanistic phases. Nature. 2011 Feb 24;470(7335):540-2. PubMed.
  2. . Prion diseases: Infectivity versus toxicity. Nature. 2011 Feb 24;470(7335):470-1. PubMed.
  3. . Stable kinesin and dynein assemblies drive the axonal transport of mammalian prion protein vesicles. Cell. 2011 Feb 18;144(4):551-65. PubMed.