Mad Cow Home or Best Links

Early fulltext release of articles by Nature:

...Jeffrey Almond, John Pattison: News and Views
...Another editorial
...Moira Bruce and strain-typing
...Collinge gel strain-typing
...Table 1: incubation periods for transmission
...Figure 1: Collinge strain scattergraph (key result) [enlarged below]

Smoking gun
Almond and Pattison on Human BSE
Implications for prion-only model
Highlights of Bruce strain-typing
Highlights of Collinge strain-typing

The 'smoking gun' of BSE

Nature editorial 2 October 1997 Volume 390 no 6650
Two studies, on pages 448 and 498 show that the strain of prion disease that causes the new variant of brain-wasting Creutzfeldt-Jacob disease (vCJD) is identical to bovine spongiform encephalopathy (BSE) in cattle. This makes it highly likely that humans with vCJD contracted it by consuming meat from cattle infected with BSE.

In experiments that have run for more than a year (see p498), Moira Bruce of the Institute of Animal Health in Edinburgh, Scotland, and her colleagues injected several different breeds of laboratory mouse with infectious brain samples from cows, patients with vCJD, patients with sporadic CJD and farmers who died of CJD after working with animals with BSE. The researchers looked critically at the incubation time (the time it took for the injected mice to fall ill), the type of brain damage caused, and the areas of brain damaged. The research shows that the histological presentation, symptoms and course of vCJD in mice is identical to that of BSE in mice, and distinct from other forms of CJD.

The strain of CJD that killed the farmers by CJD was not the same as BSE, confirming earlier results this year published elsewhere (The Lancet 350, 188; 1997).

In a second series of experiments on page 448, John Collinge of the Prion Disease Group, Imperial College School of Medicine at St Mary's, London, UK, and his colleagues describe a completely different approach that led to the same conclusion -- that vCJD is caused by the same agent as BSE. Using a biochemical assay, Collinge and colleagues show that the BSE and vCJD agents are the same, and are distinct from other forms of CJD in humans.

Collinge and colleagues show that the disease agent that causes BSE is able to 'convert' human prion protein into the highly resilient, pathogenic form of the protein, when tested in mice. Genetically engineered or 'transgenic' mice which have had their native mouse prion protein gene replaced with a human prion protein gene do, eventually, contract spongiform encephalopathy after injection with BSE contaminated material. Collinge and colleagues reported in Nature in 1995 (Nature 378, 779-783) that these mice do not become ill even 200 days after inoculation. The researchers waited longer, and now report that mice succumb after 500 days.


Nature 2 October 1997 Volume 390 no 6650
Jeffrey Almond and John Pattison
Two sets of studies, using different approaches, provide convincing evidence that the new variant of Creutzfeldt-Jakob disease is caused by the agent that is responsible for BSE in cattle. But they cannot tell us anything about the future number of cases of this variant disease.

On 20 March 1996 the UK government announced that a distinct variant of Creutzfeldt-Jakob disease (vCJD) had occurred in ten people in the United Kingdom over the previous 14 months (1). Like CJD, the variant form results in brain damage and death; but it is pathologically and clinically distinct, not least in afflicting comparatively young people. The government also stated that the most likely cause of this apparently new disease was exposure to the agent that has caused the epidemic of bovine spongiform encephalopathy (BSE).

Since then, there have been more cases of vCJD and further evidence has accumulated that supports a link between the two diseases (Table 1). But to date there has been nothing as compelling as the results described by two papers in this issue - one, by Moira Bruce and colleagues (2), based in Edinburgh, appears on page 498; the other, by a group led by John Collinge (Hill et al. (3) ), is on page 448. Both reinforce the conclusion that vCJD is distinct from other, sporadic forms of CJD (spCJD), and provide a convincing case that vCJD is caused by the same 'strain' of agent that has caused the BSE epidemic in Britain. In effect, vCJD is human BSE.

Both types of CJD, and other forms of transmissible spongiform encephalopathies (TSEs) such as scrapie, are thought to be caused by aberrant protein agents, which in turn cause the pathological modification of host-encoded prion proteins (PrPs). This is the 'protein only' hypothesis (4), although it still remains possible that some other kind of foreign macromolecule is involved. The two new studies (2), (3)are based on characterizing the strain of agent associated with vCJD and comparing it with strains from spCJD and from TSEs of other animals including BSE. The approaches differ slightly, however: although both groups use transmission potential as a defining character, Bruce et al. (2) concentrate on incubation period and pathology, whereas Hill et al. (3) focus on a particular feature - the so-called glycoform profiles - of the disease-specific PrP.

Evidence for the existence of multiple strains of TSE agent comes mainly from the Edinburgh group (5) and is based on the observation that different strains will give reproducible incubation times and pathology in certain lines of inbred mice. To define strains of TSEs, they routinely use four or five mouse lines including homozygotes for prolonged scrapie incubation periods (p7p7), for short incubation periods (s7s7) and heterozygotes (p7s7). Strains of TSEs may differ from one another both in the incubation time in a particular mouse line and in the temporal sequence in which the different lines succumb to disease. Pathological damage is measured semi-quantitatively by scoring vacuolation in nine regions of the brain and is expressed graphically as a lesion profile or 'signature'. This signature is reproducible and characteristic of a given strain (6).

Previously, the Edinburgh group characterized eight cases of BSE from different periods in the epidemic, and from different areas: all displayed highly similar strain characteristics (their phenotype), which were moreover distinct from those observed over many years in TSEs isolated from sheep, mink and a mule deer. Importantly, strain phenotype can be maintained upon transmission between species (as assessed by re-assay in indicator mice). Thus, the BSE phenotype was maintained after experimental transmission to a pig, a goat and two sheep. Identification of the same phenotype in three cases of feline spongiform encephalopathy, an outbreak of which has occurred in the UK contemporaneously with BSE (albeit at a much lower level), and in cases in kudu and nyala in British zoos, has been taken as evidence that BSE has infected these species (6).

Although their current research is not yet complete, Bruce and colleagues (2) can now also conclude the following: that the strain of agent from vCJD is: (a) the same in each of the three cases they have studied; (b) different from that seen with other forms of CJD; and (c) indistinguishable from that of BSE. The mean incubation period of vCJD in one of their inbred lines of mice (RIII) is relatively short, which is characteristic of BSE whatever its source. Moreover, the signature produced in these mice is almost superimposable on that produced by BSE from cattle or from any other species accidentally or experimentally infected with BSE. The remaining lines of mice that have been inoculated with vCJD are still under observation. Experience predicts that the C57BL line of mice will be the next to develop signs of disease, and that is now happening (M. Bruce, personal communication). Given the data so far, it seems unlikely that the final results of these lengthy experiments will give a different picture - a picture which, incidentally, confirms the view that CJD occurring in farmers in recent years is in fact the sporadic form and is not related to BSE (1), (7).

What of the work of Hill et al. (3) ? Their research focuses on the fragment size and the ratio of di-, mono- and nonglycosylated forms of the disease-specific PrP protein after treatment with protease, a protein-digesting enzyme. This glycoform profile also seems largely to be maintained upon inter-species transmission, although changes may occur in certain genotypes of recipient. Last year, using this parameter, the group obtained evidence that the vCJD agent was the same as that of BSE, although its possible similarity to other animal TSEs was not ruled out (8).

Their latest results (3) include a large number of transmissions to both transgenic mice expressing the human PrP gene (Prn-p) and their non-transgenic counterparts, and provide stronger evidence that the agent of vCJD is distinct from those of both spCJD and the iatrogenic form of CJD (a number of cases of which were caused by treatment of patients with growth hormone derived from human pituitary glands). By most criteria, vCJD and BSE are also highly similar: their glycoform profiles are indistinguishable, in both ratios and band sizes; mice suffering from the two diseases share unusual symptoms (some of the mice walk backwards); and although details of the pathology are yet to be published, the authors refer to "striking similarities" in PrP deposition patterns. In the line of inbred mice (FVB) used for this study, there are some differences in transmission potential of BSE and vCJD. This is particularly so in the human Prn-p transgenic animals, to which vCJD transmits more readily than BSE. These differences are, however, probably attributable to a 'species barrier' effect for BSE but not for vCJD.

Taken together, then, the two new sets of results complement each other and give a consistent message. But can we now draw firmer conclusions about the number of cases of vCJD that will occur in the UK? Unfortunately not. To date, there have been 21 confirmed instances in the UK (each one a tragedy in its own right, and our sympathy goes out to their families). The rate of new cases is not increasing, which provides some hope that the overall number will be relatively small, but it may take several years before we can be confident that this is not a period of comparative calm before a storm. Much depends on the average incubation time of vCJD: the longer the time, the higher the final figure is likely to be (9). At present we cannot calculate the average incubation time of BSE in humans; nor is it possible to estimate the amount of infectivity (in terms of cattle or mouse infectious doses) required to infect a human.

One observation that bears on these issues is that all of the vCJD cases examined so far are homozygous for a common amino-acid polymorphism in the human PrP protein, namely methionine at position 129. This raises the possibility that people who are homozygous for valine at this position or who are heterozygotes (about 11% and 51% of the UK population, respectively) may be relatively resistant to infection, may be subject to longer incubation times (10) or may have different symptoms. The Prn-p transgenic mice used by Hill et al. have the valine 129 version of the gene and it will be interesting to compare transmission to similar mice with methionine at that position.

Finally, these latest results (2), (3) also do not tell us anything more about the route by which the victims of vCJD were infected. There are various possibilities, including a common source for BSE and vCJD, and transmission from cattle to people through an intermediate species. But we still think that the most likely exposure was through eating beef products that included infected offal before it was banned from human food in late 1989. The report in the UK press of a case of vCJD in a vegetarian of 11 years' standing does not, we believe, invalidate this view; she may have inadvertently been exposed to contaminated beef products, or may have consumed infected beef before becoming a vegetarian.

A postscript. The UK government's decision, in March 1996, to point publicly to a probable link between BSE and vCJD was taken on the advice of the Spongiform Encephalopathy Advisory Committee (SEAC) - of which we were and are members, although here we are not writing in that capacity. At the time, the evidence connecting the two diseases was relatively slight and SEAC's advice was rightly questioned in both the scientific and popular press. Only rarely in such circumstances can science offer definitive evidence quickly, and decisions have to depend on the weighing of uncertainties. More such judgements may yet be required concerning BSE and human disease.

Jeffrey Almond is in the School of Animal and Microbial Sciences, University of Reading, Reading RG6 6AJ, UK. John Pattison is in the Medical School, University College London, Gower Street, London WC1E 6BT, UK.

1. Will, R. G. et al. Lancet 347, 921-925 (1996).
2. Bruce, M. E. et al. Nature 389, 498-501 (1997).
3. Hill, A. F. et al. Nature 389, 448-450 (1997).
4. Prusiner, S. B. Science 252, 1515-1522 (1991).
5. Bruce, M. E., McConnell, I., Fraser, H. & Dickinson, A. G. J. Gen. Virol. 72, 595-603 (1991).
6. Bruce, M. et al. Phil. Trans. R. Soc. B 343, 405-411 (1994).
7. Hill, A. F., Will, R. G., Ironside, J. & Collinge, J. Lancet 350, 188-188 (1997).
8. Collinge, J., Sidle, K. C. L., Meads, J., Ironside, J. & Hill, A. F. Nature 383, 685-690 (1996).
9. Cousens, S. N., Vynnycky, E., Zeidler, M., Will, R. G. & Smith, P. G. Nature 385, 197-198 (1997).
10. Palmer, M. S., Dryden, A. J., Hughes, J. T. & Collinge, J. Nature353, 340-342 (1991).
11. Lasmezas, C. I. et al. Nature 381, 743-744 (1996).
12. Bruce, M. E. Br. Med. Bull. 49, 822-838 (1993).

'Protein only' prions?

Nature 2 October 1997 Volume 390 no 6650
Jeffrey Almond and John Pattison
The results of Bruce et al. (2) and Hill et al. (3) underline the need to gain a better understanding of the nature of the agents responsible for transmissible spongiform encephalopathies (TSEs). The most favoured view is that the agents are composed solely of an altered form of a host-encoded protein known as PrP and lack a foreign nucleic acid; this is the 'protein only', or prion, hypothesis (4).

There are well-documented instances of TSE agents changing their characteristics, or phenotype, upon passage (repeated transmission through experimental animals). Indeed, Hill et al. report a change in fragment size upon transmission of BSE and vCJD to their transgenic mice. However, strain phenotypes can remain stable (12); thus the Edinburgh group of Bruce and colleagues have reported that different strains adapted to, and repeatedly passaged in, a single line of inbred mice retain their distinguishing features even though they must be composed of PrP protein with the same amino-acid sequence (6).

In addition, a single strain, as illustrated by BSE, can retain its phenotype even when passaged through different hosts. In these circumstances, then, strain phenotype is maintained even though the agent is composed of PrPs of different amino-acid sequences, raising the question of what determines strain phenotype - clearly, in these cases, it is not the primary structure of the PrP protein.

Supporters of the 'protein only' hypothesis argue that the small number of different strains that have been convincingly demonstrated so far can be explained by different conformations of the PrP protein, perhaps in combination with modifications such as glycosylation. Such differences of course need to 'breed true' upon passage.

On the other hand, opponents of the 'protein only' hypothesis point to the potentially large number of TSE strains (the Edinburgh group claim to have evidence for around 20) and argue that an unimaginable plasticity would be required for these to be accommodated by PrP conformation alone. They suggest that a more likely explanation is that, in addition to PrP, an informational molecular component is present in the infectious agent. Whatever the nature of the agent, our understanding of TSE biology is evidently incomplete.

Transmissions to mice indicate that 'new variant' CJD is caused by the BSE agent

Nature 2 October 1997 Volume 390 no 6650 448
M.E.Bruce, R.G.Will, J.W.Ironside, I.McConnell, D.Drummond, A.Suttie, L.McCardle,  A.Chree, J.Hope, C.Birkett, S.Cousens, H.Fraser, C.J.Bostock
There are many strains of the agents that cause transmissible spongiform encephalopathies (TSEs) or 'prion' diseases. These strains are distinguishable by their disease characteristics in experimentally infected animals, in particular the incubation periods and neuropathology they produce in panels of inbred mouse strains (1-4).

We have shown that the strain of agent from cattle affected by bovine spongiform encephalopathy (BSE) produces a characteristic pattern of disease in mice that is retained after experimental passage through a variety of intermediate species (5) ( , ) (6) (,) (7). This BSE 'signature' has also been identified in transmissions to mice of TSEs of domestic cats and two exotic species of ruminant (6) (,) (8), providing the first direct evidence for the accidental spread of a TSE between species.

Twenty cases of a clinically and pathologically atypical form of Creutzfeldt-Jakob disease (CJD), referred to as 'new variant' CJD (vCJD) (9), have been recognized in unusually young people in the United Kingdom, and a further case has been reported in France ( ) (10).

This has raised serious concerns that BSE may have spread to humans, putatively by dietary exposure. Here we report the interim results of transmissions of sporadic CJD and vCJD to mice. Our data provide evidence that the same agent strain is involved in both BSE and vCJD.

Transmissions to mice were set up from six typical sporadic cases of CJD (spCJD) and three cases of vCJD. All were homozygous for methionine at codon 129 of the 'prion protein' (PrP) gene, and none carried PrP gene mutations associated with familial disease. The spCJD cases included two dairy farmers (aged 61 and 64 years) who had had BSE in their herds and had therefore been potentially exposed to BSE-infected cattle or contaminated animal feed (11): two 'contemporary' cases (aged 55 and 57 years) with no known occupational exposure to BSE: and two 'historical' cases (aged 57 and 82 years) who had died in 1981 and 1983, before the onset of the BSE outbreak.

All of these spCJD cases were characterized by widespread spongiform vacuolation in the brain with few or no amyloid plaques. The vCJD cases (aged 29, 30 and 31 years) had clinical and neuropathological characteristics that were atypical for CJD (9). The main distinguishing neuropathological features in these and other vCJD cases are an extensive deposition of PrP amyloid in the brain as large 'florid' plaques and a prominent involvement of the cerebellum.

Panels of three inbred mouse strains (RIII, C57BL and VM) and one cross (C57BLVM) were challenged with CJD brain homogenates. Previous transmissions, using the same protocol, of BSE from eight unrelated cattle (Fig. 1b) and TSEs from three domestic cats (Fig. 1c), a greater kudu and a nyala (two exotic ruminants) have given a remarkably uniform pattern of incubation periods in these mice (5) (,) (6) (8).

The shortest incubation periods were seen in RIII mice, with means ranging from 302 to 335 days for transmissions from frozen brain samples. These isolates also produced strikingly similar patterns of vacuolar degeneration in the brains of infected mice, as represented by the 'lesion profile' (5) (,) (6) . The lesion profile is a well-established semiquantitative method of measuring the targeting of vacuolation to different brain regions, and reliably discriminates between TSE strains in mice (2). In addition, the disease characteristics in mice injected with brain from two sheep, a goat and a pig that had been experimentally infected with BSE were very similar to those seen in direct BSE transmissions from cattle (6) (7).

The BSE 'signature', based on both incubation periods and pathology, has only ever been seen in transmissions from animals suspected or known to have been infected with BSE. It has never been seen throughout an extensive series of transmissions, set up in Edinburgh between 1963 and 1994, of other naturally occurring TSEs (35 sheep and two goats with scrapie, two mink with transmissible mink encephalopathy, and a mule deer with chronic wasting disease). For example, the incubation periods and lesion profiles seen in transmissions from six sheep with scrapie, collected since 1985, are shown in Figs 1a and 2a. Within the same timescale a further two sources of sheep scrapie failed to transmit to mice. In general, natural scrapie transmissions in our own laboratory and elsewhere have given variable results, probably reflecting variation in agent strain amongst the sheep sources (12) (13).

At the time of writing, the transmissions of vCJD to mice have been in progress for 360 days. The RIII mice injected with all three vCJD sources have developed a progressive clinical disease very similar to BSE, with incubation periods in individual mice ranging from 288 to 351 days. The first signs were nervousness and hypersensitivity, followed by lethargy, weight loss, urinary incontinence and postural abnormalities.

Excluding early intercurrent deaths, all RIII mice injected with vCJD have developed disease, with mean incubation periods up to a standard clinical endpoint of 304 4, 306 6 and 310 4days (s.e.m.) for three sources (Fig. 1d), within but at the lower end of the range previously seen for BSE and related isolates (6) (Fig. 1b, c). Clinical signs are now apparent in some of the C57BL mice, an observation that is also consistent with the BSE pattern (see Fig. 1b). Diagnosis was confirmed for all clinically affected mice by the presence of vacuolar degeneration in the brain and for selected mice in all three experiments by the demonstration of relatively protease-resistant isoforms of PrP (PrP (res)) in western blots of brain extracts and pathological accumulations of PrP in immunostained brain sections.

The neuropathology in clinically affected RIII mice with vCJD was also similar to that seen in RIII mice with BSE, consisting of a mild-to-moderate grey-matter vacuolation of the hypothalamus, medulla oblongata and septum, and a more severe vacuolation of the cochlear nucleus. Amyloid plaques were not a prominent feature of this pathology. The lesion profiles in RIII mice for the three sources (Fig. 2d) were very similar to each other and also to those in transmissions to RIII mice of BSE (Fig. 2b), TSEs of cats and exotic ruminants (Fig. 2c), and experimental sheep, goat and pig BSE (6) (,) (7), but differed markedly from those seen in transmissions from sheep with natural scrapie (Fig. 2a). Although results are so far only available for the RIII mouse strain, the striking similarity between vCJD and BSE in these mice, in terms of both incubation periods and pathology, is in itself evidence that the same strain of agent is involved in vCJD and BSE.

In contrast to the results with the vCJD sources, no clinical signs of neurological disease have yet been seen in any mice in the six transmissions of spCJD, although they have been in progress for between 600 and 800 days. Figure 3 shows survival curves for RIII mice in these experiments, compared with survival curves in BSE, feline spongiform encephalopathy (FSE) and vCJD transmissions. No significant differences in median survival times were found between RIII groups challenged with the six spCJD sources, or between these groups and saline-injected controls.

However, this does not indicate a failure to transmit spCJD, as vacuolar degeneration typical of TSE infection was seen in the brains of some mice dying with intercurrent disease in all six experiments, from about 400 days after challenge. This pathology has so far been seen in 130 of the 156 mice surviving beyond 500 days after injection for which brain was available for histopathological scrutiny.

No such changes have been seen in the control mice of any age in this set of experiments, or in mice of the same strains injected with human brain homogenates from patients with amyotrophic lateral sclerosis or laryngeal carcinoma in a previous study (14). Western blot and immunohistochemical analyses have demonstrated the accumulation of PrP (res) in selected brains showing vacuolar pathology, confirming successful transmission of a TSE from all six spCJD sources.


A series of transmissions to mice of spCJD and familial human TSEs associated with mutations in the PrP gene have been reported in Japan ( ) (15). Although different mouse strains were used in the Japanese series, the results for transmissions of spCJD from 129-methionine sources were broadly similar to ours in that transmission was achieved from all sources and mean incubation periods in recipient mice were long (573-863 days) (15).

Transmissions of the familial TSE Gerstmann-Straussler-Scheinker syndrome (GSS) were achieved from only one-third of the sources tested, but the mean incubation periods in successful transmissions were relatively short (237-517 days) (15). Although some of these incubation periods were quite close to our results for vCJD in RIII mice, the pathology in mice with GSS was strikingly different as it included a prominent vacuolation of white-matter tracts (16). The Japanese workers also reported the transmission of another human familial TSE, fatal familial insomnia (FFI), with a mean incubation period of 455 days in recipient mice (17). The pathology in mice with FFI was indistinguishable from that in mice with spCJD in the Japanese series, apart from there being a more pronounced involvement of the thalamus.

Our results highlight several fundamental features of the TSEs previously established using experimental isolates ( ) (3) (,) (4). The consistency in transmission properties shows that the agent must interact with genetic factors in the host to control the timing and neuropathology of the disease with extraordinary precision. Different strains of agent (spCJD, vCJD) can be isolated from hosts with the same PrP amino-acid sequence (in this case, patients with the 129-methionine genotype) but, conversely, the same strain of agent can be detected in hosts with different PrP sequences (so far the BSE 'signature' has been seen in transmissions from eight different species (6)). This clearly indicates that TSE agents carry some form of information that specifies strain-specific properties, but the molecular basis of this information is still a matter for speculation (4).

It has been reported that vCJD can be distinguished from spCJD by the relative prominence of differently glycosylated forms of PrP (res) and the molecular size of the unglycosylated form (18). Samples from dairy farmers with CJD have given glycoform ratios resembling those from other cases of spCJD (19). A similarity in glycoform patterns between vCJD and BSE has been presented as evidence of a link between the two ( ) (18). However, a 'BSE-like' glycoform pattern has also been seen for experimental scrapie isolates that are unrelated to BSE (20) and for FFI in humans (21). Therefore, although the analysis of PrP diversity provides a useful supplement to strain typing in mice, it is premature to draw conclusions concerning causative links between TSEs in different species on the basis of glycoform-ratio analysis alone. A full analysis of glycoform patterns in the present series of transmissions will be reported in due course.

In conclusion, strain typing based on transmission to mice has shown: that vCJD is caused by the same strain of agent that has caused BSE, FSE and TSEs in exotic ruminants: that vCJD is distinguishable from spCJD: and that CJD in two dairy farmers is of the spCJD type and is not linked to the causative agent of BSE. Epidemiological surveillance continues to indicate that vCJD is a new condition occurring almost exclusively in the UK. Our transmission studies, in combination with the surveillance data, provide compelling evidence of a link between BSE and vCJD.

1. Dickinson, A. G. & Meikle, V. M. H. Host-genotype and agent effects in scrapie incubation: change in allelic interaction with different strains of agent. Mol. Gen. Genet. 112, 73-79 (1971).
2. Fraser, H. & Dickinson, A. G. Scrapie in mice: agent-strain differences in the distribution and intensity of grey matter vacuolation. J. Comp. Pathol. 83, 29-40 (1973).
3. Bruce, M. E., McConnell, I., Fraser, H. & Dickinson, A. G. The disease characteristics of different strains of scrapie in Sinc congenic mouse lines: implications for the nature of the agent and host control of pathogenesis. J. Gen. Virol. 72, 595-603 (1991).
4. Bruce, M. E. Scrapie strain variation and mutation. Br. Med. Bull. 49, 822-838 (1993).
5. Fraser, H., Bruce, M. E., Chree, A., McConnell, I. & Wells, G. A. Transmission of bovine spongiform encephalopathy and scrapie to mice. J. Gen. Virol. 73, 1891-1897 (1992).
6. Bruce, M. et al. Transmission of bovine spongiform encephalopathy and scrapie to mice: strain variation and the species barrier. Phil. Trans. R. Soc. Lond. B 343, 405-411 (1994).
7. Foster, J. D., Bruce, M., McConnell, I., Chree, A. & Fraser, H. Detection of BSE infectivity in brain and spleen of experimentally infected sheep. Vet. Rec. 138, 546-548 (1996).
8. Fraser, H. et al. Transmission of feline spongiform encephalopathy to mice. Vet. Rec. 134, 449 (1994).
9. Will, R. G. et al. A new variant of Creutzfeldt-Jakob disease in the UK. Lancet 347, 921-925 (1996).
10. Chazot, G. et al. New variant of Creutzfeldt-Jakob disease in a 26-year-old French man. Lancet 347, 1181 (1996).
11. Cousens, S. N. et al. Sporadic Creutzfeldt-Jakob disease in the United Kingdom: epidemiological data from 1970-1996. Br. Med. J. 315, 389-396 (1997).
12. Dickinson, A. G. in Slow Virus Diseases of Animals and Man (ed. Kimberlin, R. H.) 209-241 (North-Holland, Amsterdam, 1976).
13. Carp, R I. & Callahan, S. M. Variation in the characteristics of 10 mouse-passaged scrapie lines derived from five scrapie-positive sheep. J. Gen. Virol. 72, 293-298 (1991).
14. Fraser, H., Behan, W., Chree, A., Crossland, G. & Behan, P. Mouse inoculation studies reveal no transmissible agent in amyotrophic lateral sclerosis. Brain Pathol. 6, 89-99 (1996).
15. Tateishi, J. Transmission of human prion diseases of rodents. Semin. Virol. 7, 175-180 (1996).
16. Tateishi, J., Ohta, M., Koga, M., Sato, Y. & Kuroiwa, Y. Transmission of chronic spongiform encephalopathy with kuru plaques from humans to small rodents. Ann. Neurol. 5, 581-584 (1979).
17.Tateishi, J. et al. First experimental transmission of fatal familial insomnia. Nature 376, 434-435 (1995).
18. Collinge, J., Sidle, K. C. L., Meads, J., Ironside, J. & Hill, A. F. Molecular analysis of prion strain variation and the aetiology of 'new variant' CJD. Nature 383, 685-690 (1996).
19. Hill, A. F., Will, R. G., Ironside, J. & Collinge, J. Type of prion protein in UK farmers with Creutzfeldt-Jakob disease. Lancet 350, 188 (1997).
20. Somerville, R. A. et al. Biochemical typing of scrapie strains. Nature 386, 564 (1997).
21. Telling, G. et al. Evidence for the conformation of the pathological isoform of the prion protein enciphering and propagating prion diversity. Science 274, 2079-2082 (1996).
22. Dickinson, A. G., Meikle, V. M. H. & Fraser, H. Identification of a gene which controls the incubation period of some strains of scrapie agent in mice. J. Comp. Pathol. 78, 293-299 (1968).
23. Westaway, D. et al. Distinct prion proteins in short and long scrapie incubation period mice. Cell 51, 651-662 (1987).
24. Fraser, H. & Dickinson, A. G. The sequential development of the brain lesions of scrapie in three strains of mice. J. Comp. Pathol. 78, 301-311 (1968).
25. Farquhar, C. F. et al. in Transmissible Spongiform Encephalopathies (eds Bradley, R. & Marchant, B.) 301-313 (Commission of the European Communities, Brussels, 1994).
26. Bell, J. E. et al. Prion protein immunocytochemistry - UK five centre consensus report. Neuropathol. Appl. Neurobiol. 23, 26-35 (1997).
27. Kirkwood, B. R. Essentials of Medical Statistics (Blackwell, Oxford, 1988).
28. Mardia, K. V., Kent, J. T. & Bibby, J. M. Multivariate Analysis (Academic, London, 1979).

We acknowledge the contribution of A. Dickinson, who pioneered TSE strain discrimiantion in the 1960s.

The same prion strain causes vCJD and BSE

AF.Hill, M Desbruslais S Joiner, Katie Sidle Ian Gowland, John Collinge LJ. Doey, Peter Lantos
Epidemiological and clinicopathological studies, allied with pathological prion protein (PrP(Sc)) analysis, strongly support the hypothesis that the human prion disease new variant Creutzfeldt-Jakob disease (vCJD) is causally related to bovine spongiform encephalopathy (BSE)(1),(2), but considerable controversy remains. Distinct prion strains are distinguished by their biological properties on transmission to laboratory animals and by physical and chemical differences in PrPSc strains. We now find that the biological and molecular transmission characteristics of vCJD are consistent with it being the human counterpart of BSE.

We studied transgenic mice expressing only human PrP (HuPrP(+/+) Prn-p(o/o)), which have been shown to lack a species barrier to human prions from one iatrogenic CJD case(3), comparing them with non-transgenic (FVB) mice. All of 16 further CJD cases, encompassing a wide range of clinicopathological phenotypes, all three PrP(Sc )types reported in sporadic and acquired prion diseases(2) and all PRNP genotypes at polymorphic codon 129, a key determinant of genetic susceptibility to human prion diseases(4-6), were transmitted to these transgenic mice.

Almost all inoculated transgenic mice contracted disease with similar short incubation periods, consistent with a lack of species barrier to these isolates (Table 1). These transgenic mice express human PrP homozygous for valine at codon 129. However, there was no significant difference in mean incubation periods between inocula of the different codon 129 genotypes. PrP(Sc) typing of these transmissions showed that the same prion types seen in sporadic and iatrogenic CJD (types 1-3) are produced, distinct from that seen in vCJD (type 4)(2). Only occasional transmissions, at longer and variable incubation periods, were seen in FVB mice.

In contrast, efficient transmission of vCJD to FVB mice was observed (Table 1) although incubation periods were prolonged. Conversely, the attack rate of vCJD in the transgenic mice was reduced in comparison to typical CJD, and incubation periods were generally more variable and prolonged. Mean incubation periods to these six vCJD cases were similar in both types of mice. The clinical course in vCJD-inoculated transgenic mice was much longer than in transmissions of typical CJD. vCJD in humans is also associated with a long clinical duration(1). Some mice, as well as showing typical neurological features, persistently walked backwards. This unusual clinical sign was not seen in transmissions of typical CJD, fatal familial insomnia or other inherited prion diseases(7).

BSE transmits efficiently to FVB mice(3), albeit with prolonged and variable incubation periods (Table 1) which fall to a consistent short incubation period of around 140 days on second passage (data not shown). Transmissions of BSE into the transgenic mice did not occur at incubation periods well beyond those of classical CJD(3), but we have now observed transmission with much longer incubation periods (Table 1). These transmissions resembled those of vCJD with a long clinical duration and backwards walking in some animals as well as the otherwise typical clinical features of mouse scrapie.

There were striking similarities in PrP deposition patterns between BSE- and vCJD-inoculated animals (detailed neuropathological studies will be published elsewhere). Such patterns are determined by host genotype as well as by agent strain(8). We saw distinct patterns in the two types of mice, but, in each case, vCJD and BSE produced closely similar patterns. In vCJD- and BSE-inoculated non-transgenic mice, there were PrP plaques and diffuse PrP deposition. In vCJD- and BSE-inoculated HuPrP(+/+) Prn-p(o/o) transgenic mice we saw a predominantly pericellular pattern of PrP immunostaining (data not shown). PrP plaques are a rare feature of prion disease in mice. Occasional mock-inoculated transgenic mice showed weaker and less extensive pericellular PrP immunostaining, probably reflecting the high level of PrP(C) overexpression in these mice. Western blotting for PrP(Sc) was negative in all these controls.

We performed western blot analysis to determine the PrP(Sc) types produced in these transmissions. We have previously shown that the PrP(Sc) type seen in vCJD (type 4) has a ratio of glycoforms closely similar to that of BSE passaged in several other species(2). vCJD-inoculated FVB mice produced mouse PrP(Sc) with type 4-like glycoform ratios and fragment sizes indistinguishable from those in BSE-inoculated FVB mice (Fig. 1a,b).

In transmission of vCJD to HuPrP(+/+) Prn-p(o/o) transgenic mice, where human PrP(Sc) is generated, fragment sizes in inoculum and host can be directly compared. Again the PrP(Sc) produced had type 4-like glycoform ratios. However, the fragment sizes differ from those in the inoculum and were indistinguishable from those in the type-2 PrP(Sc) pattern(2) (Fig. 1c). We have designated this new pattern type 5.

A change of fragment size on passage in mice of a different codon 129 PrP genotype than the inoculum has been reported previously(2). Type-1 PrP(Sc), seen in CJD cases of 129MM PRNP genotype, consistently converts to type-2 PrP(Sc) on passage in these transgenic mice expressing 129VV human PrP. The glycoform ratios of the original inoculum are also maintained(2). Abrupt changes in the biological properties ('mutation') of murine scrapie strains on passage in mice of different genotypes are well recognized(9). We have not, however, been able to show PrP(Sc) by western blotting in BSE- inoculated HuPrP(+/+) Prn-p(o/o) transgenic mice. This may reflect culling of many of these mice soon after clinical diagnosis rather than at a more advanced clinical stage. Though transmission of prion diseases without detectable PrP(Sc) on primary passage has been reported(7),(10), it will be important to confirm transmission by second passage studies.

The prion titres in these primary inocula are unknown but may be higher in the human cases, because cattle with BSE will have been culled before the terminal stages of disease. However, on clinical, pathological and molecular criteria, vCJD shows remarkable similarity in its transmission characteristics to BSE, and is quite distinct from all other forms of sporadic and acquired CJD. These data provide compelling evidence that BSE and vCJD are caused by the same prion strain. Taken together with the temporal and spatial association of vCJD with BSE but not with scrapie or other animal prion diseases, and BSE transmission studies in macaques(11), this strongly suggests that vCJD is caused by BSE exposure. The theoretical possibility that both BSE and vCJD arise from exposure to a common unidentified source appears remote.

The production of a distinct molecular strain type on transmission of vCJD to mice expressing valine 129 human PrP suggests that BSE transmitted to humans of this genotype might produce a similar strain. Such cases may differ in their clinical and pathological phenotype to vCJD, but could be identified by PrP(Sc) typing.

Although it has been argued that the species barrier resides in PrP primary structure differences between donor and host(12), our data emphasize that strain type can be as important. As prion propagation involves interactions between PrP(Sc) and host PrP(C), and strains are associated with differences in PrP conformation and glycosylation(2),(13), such PrP interactions may be most efficient if the interacting proteins are not only of the same sequence but have similar conformational preferences and glycosylation. Mismatch of codon 129 between inoculum and HuPrP(+/+) Prn- p(o/o) mice does not significantly affect CJD transmission, but this could differ for BSE. All vCJD cases have been 129MM genotype (ref. 14 and unpublished data). Although our 129VV mice are much less susceptible to BSE than to typical CJD, suggesting a substantial species barrier, 129MM human PrP mice could be more susceptible.

1.Will, R. G. et al. Lancet 347, 921-925 (1996).
2.Collinge, J., Sidle, K. C. L., Meads, J., Ironside, J. & Hill, A. F. Nature 383, 685-690 (1996).
3.Collinge, J. et al. Nature 378, 779-783 (1995) ; addendum 389, 526 (1997).
4.Baker, H. F. et al. Lancet 337, 1286 (1991).
5.Collinge, J., Palmer, M. S. & Dryden, A. J. Lancet 337, 1441-1442 (1991).
6.Palmer, M. S., Dryden, A. J., Hughes, J. T. & Collinge, J. Nature 352, 340-342 (1991).
7.Collinge, J. et al. Lancet 346, 569-570 (1995).
8.Bruce, M. E. et al. J Gen. Virol. 72, 595-603 (1991).
9.Bruce, M. E. Br. Med. Bull. 49, 822-838 (1993).
10.Lasmézas, C. I. et al. Science 275, 402-405 (1997).
11.Lasmézas, C. I. et al. Nature 381, 743-744 (1996).
12.Prusiner, S. B. et al. Cell 63, 673-686 (1990).
13.Telling, G. C. et al. Science 274, 2079-2082 (1996).
14.Collinge, J. et al. Lancet 348, 56 (1996).
15.Piccardo, P. et al. J. Neuropathol. Exp. Neurol. 56, 589 (1997).

Mad Cow Home or Best Links