Creutzfeldt–Jakob disease, Gerstmann–Straussler–Scheinker (GSS) syndrome, scrapie and bovine spongiform encephalopathy are part of a family of neurodegenerative diseases caused by changes in the metabolism of a plasma-membrane protein, PrPC or prion protein. These prion diseases can be acquired by infection, arise sporadically by an unknown mechanism, or result from inheritance of a defective PrP gene. An abnormal PrP (PrPSc) is characteristic of most prion diseases, but in some inherited cases the PrP that accumulates in the brain is not PrPSc but CtmPrP, a transmembrane form of the protein1. And on page 822 of this issue, Hegde and co-workers2 show that CtmPrP may be a neurotoxic molecule that is common to both the genetic and the acquired forms of prion diseases.
The prion hypothesis holds that an abnormal prion protein (PrPSc) may be the sole component of the ‘transmissible particle’ that causes acquired forms of disease. But there are several anomalies. For example, high levels of a GSS-syndrome mutant PrP (P101L) in the brains of transgenic mice (the Tg[MoPrP(P101L)] strain) induced a spontaneous neurological disorder in the absence of conventional PrPSc, although this disorder could be transmitted to a limited range of host rodents3. A subset of people with GSS syndrome has been found with a different mutation (A117V) in the PrP gene that also does not produce conventional PrPSc. However, this form of the disease cannot be transmitted to laboratory rodents4. Further anomalies in the prion hypothesis have been exposed by cross-species transmission experiments5,6, in which recipient mice develop disease without the formation of recognizable PrPSc.
These anomalies have led some to reject the prion hypothesis in favour of a more conventional viral agent as the cause of transmissible spongiform encephalopathies. Hegde and co-workers2, by contrast, have looked for a solution to these difficulties by understanding the effect of mutations on the cell-membrane topography and stability of PrPC, in vitro and in vivo. Normally, PrPC is fully translocated across the membrane of the endoplasmic reticulum, then anchors to its lumenal face before being processed and transported to the cell surface. But sometimes it can get trapped in the membrane, and it forms the CtmPrP form. The carboxy- terminal (Ctm) segment of this protease-sensitive intermediate pokes out into the lumen of the endoplasmic reticulum, whereas its amino-terminal domain is stuck in the cytosol on the other side — like a breech baby, it seems to be born tail first.
In a previous publication1, Hegde and colleagues found a correlation between the propensity of mutants to form CtmPrP and their ability to induce neurodegenerative disease (in a dose-related manner) in transgenic mice. Elevated levels of CtmPrP were also associated with a case of GSS syndrome (A117V), and CtmPrP was suggested to be involved in the brain degeneration associated with non-transmissible, ‘genetic’ prion diseases. Hegde et al.2 have now tested whether this CtmPrP-associated disease can be transmitted. To do this, they took brain homogenates from one line of transgenic mice, known as [Tg[SHaPrP(KH→II)H], that express high levels of a hamster CtmPrP. These mice rapidly develop spontaneous disease, characterized by high levels of CtmPrP but no detectable PrPSc. The authors then inoculated a panel of transgenic mice and hamsters with these brain homogenates. They observed no significant differences in survival times between animals inoculated with Tg[SHaPrP(KH→II)H] mouse brain and those inoculated with control tissues. In other words, CtmPrP-associated prion disease is not transmissible.
The hamster PrP protein expressed in Tg[SHaPrP(KH→II)H] mice has amino-acid changes at codons 109 (where lysine is replaced by isoleucine) and 110 (histidine by isoleucine). Hegde et al. next identified other mutations leading to an in vitro increase in levels of CtmPrP, and produced lines of transgenic mice expressing high and low levels of the equivalent PrP proteins. High expressors invariably developed spontaneous disease, with CtmPrP but no protease-resistant PrPSc in their brains. Low expressors were clinically normal, and did not show accumulation of CtmPrP. The authors defined the product of the expression levels in Tg[SHaPrP (KH→II)H] mice, and the ratio of CtmPrP to total PrP formed in a coupled transcription–translation system, as the ‘Ctm-index’ — a semi-quantitative measure of the propensity of a transgenic line to generate CtmPrP. They then used their transgenic mice to see whether there is a relationship between CtmPrP and PrPSc during acquired disease.
When Hegde and co-workers infected their mouse lines with a hamster scrapie strain (Sc237), they found that the amount of PrPSc accumulating at the onset of clinical disease seemed to be inversely related to the Ctm-index. This was the case irrespective of whether comparisons were made between mice with constant levels of PrP expression but different mutations, or between mice with the same mutation and different levels of expression. Amino-terminally truncated PrP (ref. 7) or Dpl (a PrP-like molecule)8, both of which lack an amino-terminal repetitive sequence, have been linked to neurodegenerative diseases, and CtmPrP may induce cell death by a similar mechanism (Fig. 1).
All the available data fit with a model for an interrelationship of CtmPrP and PrPSc, in which CtmPrP causes the neurodegeneration seen in both genetic and transmissible prion protein diseases. One prediction of the model is that the accumulation of PrPSc should increase the generation of CtmPrP. The authors used an ingenious approach to test this prediction and solve the problem of discerning PrPSc from CtmPrP in infected transgenic mice. They produced a mouse line that makes both mouse PrP (MoPrP) and hamster PrP (SHaPrP), and infected that line with the RML strain of scrapie. This strain can convert only mouse PrP — not hamster PrP — to PrPSc. Using a hamster-specific monoclonal antibody to distinguish between CtmSHaPrP and MoPrPSc, they showed a positive correlation between the amounts of PrPSc and CtmPrP accumulating in the brain during infection.
When some prion diseases are transmitted between species, PrPSc is not observed. Could the existence of a neurotoxic CtmPrP resolve this anomaly? Yes if one assumes that, when transmitting across species barriers (such as from cow to mouse5,6 or from human to Tg[MoPrP(P101L)] mouse6), the balance between CtmMoPrP and MoPrPSc production is heavily in favour of CtmMoPrP. This would lead to a transmitted, neurological disease provoked by CtmPrP accumulation in the virtual absence of PrPSc. Similarly, if Tg[MoPrP(P101L)] mice3 have a high Ctm-index, the anomaly of their spontaneous disease in the absence of PrPSc could be explained by the preferential accumulation of CtmPrP. However, it then becomes an even greater mystery as to why this MoPrP(P101L) ‘spontaneous’ disease should be transmissible whereas the other PrP spontaneous diseases are not.
Hegde, R. S. et al. Science 279, 827–834 (1998).
Hegde, R. S. et al. Nature 402, 822–826 (1999).
Hsiao, K. K. et al. Proc. Natl Acad. Sci. USA 91, 9126–9130 (1994).
Tateishi, J. et al. Neurology 40, 1578–1581 (1990).
Lasmezas, C. I. et al. Science 275, 402–405 (1997).
Manson, J. et al. EMBO J. 18, 6855–6864 (1999).
Shmerling, D. et al. Cell 93, 203–214 (1998).
Moore, R. C. et al. J. Mol. Biol. 292, 797–817 (1999).
Brown, D. R. et al. Nature 390, 684–687 (1997).
Rieger, R. et al. Nature Med. 3, 1383–1388 (1997).
Caughey, B. et al. J. Virol. 68, 2135–2141 (1994).
Harris, D. A. et al. Curr. Top. Microbiol. Immunol. 207, 77–93 (1996).
About this article
Mechanisms of prion-induced modifications in membrane transport properties: Implications for signal transduction and neurotoxicity
Chemico-Biological Interactions (2001)
Nature Reviews Molecular Cell Biology (2001)
Journal of Neuroscience Research (2000)
Cell Death & Differentiation (2000)
The Veterinary Journal (2000)