In search of elusive fragments of death

BSE! CJD! Fin de siecle acronyms! Over 30 young people have died horrible deaths from the latter, and maybe many more will follow. More than £3 billion has been spent by the Government on propping up beef. And yet never in recent times have such expensive policy decisions rested on such scientific uncertainty. Why? It is not the scientists' fault. It is because of the appallingly difficult science.

For decades it has been known that the causative agents of diseases like BSE, CJD and scrapie - the transmissible spongiform encephalopathies (TSEs) - are very different from the viruses and bacteria that we have been successfully isolating and characterising for a hundred years.

A good way to find out about infectious agents is to challenge them with chemical and physical insults and measure the decline in infectivity. Not only does this tell us a lot about their makeup and structure, but it provides essential information about how to control them with disinfectants and sterilising procedures. Viruses and bacteria (except spore-formers) are rapidly killed by boiling, for example, and succumb pretty speedily to treatment with reactive chemicals. Not so the TSEs. They die very slowly under heat treatment and relish formaldehyde. So they are not ordinary bacteria or viruses. How big are they? Radiation treatment (the smaller a living object, the smaller target it presents and the harder it is to kill) suggests they are very small - in DNA terms a lot smaller than the average gene. And the diseases that they cause are unlike straightforward infections because they can have enormously long incubation periods. The illnesses themselves develop very slowly - and are uniformly fatal. No protective immune response occurs. Not only is this bad news for patients, but it means that one extremely important diagnostic method - measuring antibody levels - is missing.

The very slow infectious process and the absence of immunological handles are major obstacles to laboratory work on TSEs. Nevertheless their unusual properties intrigue scientists so much - Francis Crick has even speculated that they present a challenge to his central dogma of molecular biology - that enormous efforts have been made to find out what they are. Closure of the field is not at hand, however - despite the award of two Nobel prizes.

The essential problem is that the agent has not been purified. That is to say, no one has yet produced a test-tube that contains the molecular entity that causes the disease - a protein or nucleic acid or a defined complex of the two - and nothing else.

Fatal Protein, like all the other rapidly increasing number of books on TSEs, has therefore to be looked on as an account of "work in progress". Is it successful? Up to a point. But beware. It is very partisan. Although avowedly aimed at the general reader, its account of a central issue - the nature of the agent - is unbalanced. To be fair, its title gives warning, and the authors honestly admit that they "concentrate on the 'protein only' hypothesis" and leave other views to "another book, written by someone who supports them". But for a popular account this will not do. Although there are good and comprehensive chapters on diseases like scrapie, Kuru, CJD, BSE and new-variant CJD, the authors' hostility to any scientist or any observation that does not support the line that the infectious agent is a protein, the prion protein, goes too far. Stan Prusiner, its chief proponent ("most distinguished") publishes his "revolutionary ideas" in "prestigious journals". Alan Dickinson and his Edinburgh colleagues Moira Bruce and Hugh Fraser, whose virtuoso genetics and painstaking strain typing laid the basis for most recent advances in TSE work, and provide some of the best evidence of a link between BSE and new-variant CJD, are rubbished as polemicists who go over the top or as scientists who do not know how to interpret their results properly.

Is this because their demonstration of many strain types mounts a major challenge to the Prusiner hypothesis, and lends support to those who postulate that the agent contains another informational molecule - maybe a small nucleic acid? I suspect so. So it ill behoves the authors to lecture us on "the scientific method being not quite so high minded or rational as we might be led to believe". Read this book, but with care. It has a lot of useful information in it. Those with a biological background will profit from it. But also read other books - such as Richard Rhodes's Deadly Feasts - as an antidote to its bias.

Hugh Pennington is professor of medical bacteriology, University of Aberdeen.