BSE and the prion wars

The debate over what causes diseases such as CJD heats up again: an article in `Science' revives an old theory; meanwhile there have been another two deaths. Charles Arthur reports

It was one of those phone calls that are rather hard to make. "Hello? Is that Professor Laura Manuelidis? I'm calling from The Independent in London, about the paper you authored in Science this week - the one about BSE and CJD, and the experiments in mice." "Yes?" said Prof Manuelidis. "Well. Before I read it, I thought I understood the science of BSE and CJD, at least a bit."

"Yes?"

"But I can't understand your paper at all."

It came as some relief that rather than sighing, Professor Manuelidis laughed. But her reply then pointed towards a schism which is still nagging away at scientists, and everyone else who is trying to understand these deadly diseases.

"Of course you can't," she said. "You've been stuck in the prion mind- set for so long."

The "prion mind-set"? Certainly, there is such a thing. It is the dominant explanation at present among many scientists (and journalists) studying "mad cow disease" - bovine spongiform encephalopathy, or BSE - and the related diseases, such as Creutzfeldt-Jakob Disease (CJD) in humans and scrapie in sheep. It says that this group of diseases, known as the transmissible spongiform encephalopathies (TSEs), are caused by a spontaneous change in a cell protein called PrP. It changes from a soluble form, which the body manufactures, into an insoluble one that is useless to the cell. The prevailing theory is that this misshapen PrP protein - perhaps aided by some other protein or process - triggers others to change shape. Eventually the effect causes an avalanche. The insoluble residue of the protein is found in the brains of victims of the TSEs, along with the spongy "holes" that characterise these irreversible, fatal illnesses.

The misshapen, trigger protein is called a "prion" - a term co-opted by Stanley Prusiner of the University of California at San Francisco in the early Eighties. His "prion" theory has gradually won supporters on both sides of the Atlantic.

But detractors remain determined to show that the TSEs are not caused by lone proteins at all, but by living things - viruses, or similar DNA or RNA packages. This is the "slow virus" or "virino" hypothesis, which held sway before Prusiner. Its backers say it is axiomatic that an infectious illness can be caused only by something that contains DNA or RNA, and so can replicate itself. To them, the idea of a protein conformation replicating itself seems absurd.

Professor Manuelidis thinks that the prion hypothesis is wrong: that PrP is important, but it's not the whole story.

Does that matter? Certainly: if these fatal diseases are caused, or triggered, by a virus-like entity, it may be possible to arouse an immune response to them and prevent them taking hold. And it does matter: yesterday the Department of Health officially released figures showing that in May two more Britons died of the "new variant" of CJD - bringing the total number of victims to 20 (19 in the UK and one in France). The Government's advisory committee Seac has summarised the research since March 1996, and deemed that it is "consistent with the hypothesis" that v-CJD is "caused by exposure to the BSE agent, and no evidence refuting that evidence has yet come to light." BSE infected about 1 million British cows.

The problem is that we still don't know exactly what the agent is. Nobody has ever managed to isolate it. The pro-prion group have never produced normal PrP in a test tube, changed it to malformed PrP, then shown that that alone is sufficient to cause a TSE. But the anti-prion group (who usually propose the idea of a "slow virus", which takes up to 30 years to take effect) have problems explaining the way that processes which should destroy DNA or RNA - such as intense radiation - don't reduce the infectiousness of TSE samples.

As it happens, both sides are represented on Seac. But the paper written by Professor Manuelidis and two colleagues from the Yale School of Medicine and published in last Friday's Science appears to have important findings for the anti-prion group. Perhaps most significant was the discovery of an immune response early in the disease cycle in the spleen of rats infected with extracts from human CJD victims.

Why is that important? "We found an inflammatory response, but no insoluble PrP," says Prof Manuelidis. "The animals were clearly dying and sick." If the infection were just caused by an internal cell protein, there should be no immune response. It's an unusual result; though it echoes another paper published earlier this year in Science in which a team of French researchers were able to pass TSEs between mice which were not showing the insoluble form of the protein." Prof Manuelidis adds: "This is an infectious disease. I think it's a virus. Most viruses can have variants which make them more virulent; the BSE cows were infected by a very virulent form." Passing CJD to subsequent generations of rats - "passaging" it - produced more virulent forms. She also found that the level of insoluble PrP did not directly match the number of spongy holes found in the brain: "Almost all host PrP [the cellular version] was suddenly altered at a later stage of the disease. In contrast, when vacuolisation [the spongy holes] was obvious after 150 days, [insoluble] PrP accounted for less than 1 per cent of that found at 300 days." The most likely explanation, she suggests, is that the new, more virulent infectious agent did its work of destroying the brain "by non-PrP pathways".

That's an important finding, if repeated, since it challenges the prion hypothesis head-on. It simply wouldn't make sense for the prion to change its host protein. Prof Manuelidis suggests that, instead, there is some initial infection which then leads - over a very long time - to the viral effect, and subsequently to the build-up of insoluble PrP. The PrP is a symptom, not a cause or process, by that scenario.

Might that be right? Jeff Almond, professor of virology at the University of Reading, and a prominent member of Seac, is dubious of the idea that "passaging" would automatically make something which is already infectious more virulent. He does not belong strongly to either the prion or the anti-prion camp, preferring first principles. "What I teach people here at the university is that evolution is the important thing. The evolutionary pressure is largely downwards, in terms of virulence. If a virus kills all its hosts before they can pass it on, it doesn't survive. Even smallpox and the bubonic plague only killed about 30 to 40 per cent of the people who got it. No pathogen can survive if it kills 100 per cent of its hosts." Also, he adds, "We're also not seeing increasing virulence in the strains of TSE in cows or sheep."

But might that not be because the BSE agent is already remarkably virulent - as shown by the fact that it can infect so many different species, such as cats, kudu, pigs, and a laboratory lexicon - possibly including humans?

"Well, it comes down to this: there's no evidence that any form of CJD is passed to people who live with victims." Perhaps not, but the idea of a virus - in some form - which might take years to take effect, and whose virulence depends on the host's genotype, isn't so far-fetched. And if it takes between 10 and 30 years to kill its host, isn't that long enough for most host species to have produced more children - guaranteeing that there will be more to infect?

"It's a good question," said Almond. "It's an interesting question."

And so is the question of whether the prion theory can explain Prof Manuelidis's work. Her response? "Oh, of course it will. Someone always finds a way to twist it all round."

It may seem like more turf wars - on a par with arguments about how life started, or the universe. But while these wars are being played out in the science magazines, people are still dying - a fact of which the scientists are acutely aware.