both human and mouse prion genes will be slower to get human C J D than a mouse with only a human prion gene: does this mean different prions compete? We do not know.
How does the gene change its strain as it moves through a new species? Mice cannnot easily catch hamster scrapie, but when they do, they pass it on with progressively greater ease to other mice.3
Why? We do not know.
Why does the disease spread from the site of injection slowly and progressively, as if bad prions can only convert good prions in their immediate vicinity? We know the disease moves through the B cells of the immune system, which somehow transmit it to the brain.4
But why them, and how? We do not know.
The truly baffling aspect of this proliferating knowledge of ignorance is that it strikes at the heart of an even more central genetic dogma than Francis Crick's. It undermines one of the messages I have been evangelising since the very first chapter of this book, that the core of biology is digital. Here, in the prion gene, we have respectable digital changes, substituting one word for another, yet causing changes that cannot be wholly predicted without other knowledge. The prion system is analogue, not digital. It is a change not of sequence but of shape and it depends on dose, location and whether the wind is in the west. That is not to say it lacks determinism. If anything, C J D is even more precise than Huntington's disease in the age at which it strikes. The record includes cases of siblings who caught it at exactly the same age despite living apart all their lives.
Prion diseases are caused by a sort of chain reaction in which one prion converts its neighbour to its own shape and they each then convert another, and so on, exponentially. It is just like the fateful image that Leo Szilard conjured in his brain one day in 1933, while waiting to cross a street in London: the image of an atom splitting and emitting two neutrons, which caused another atom to split and emit two neutrons, and so on — the image of the chain reaction that later exploded over Hiroshima. The prion chain reaction is of course much slower than the neutron one. But it is just P O L I T I C S 2 8 1
as capable of exponential explosion; the New Guinea kuru epidemic stood as proof of this possibility even as Prusiner began to tease out the details in the early 1980s. Yet already, much closer to home an even bigger epidemic of prion disease was just starting its chain reaction. This time the victims were cows.
Nobody knows exactly when, where or how — that damned mystery again — but at some time in the late 1970s or early 1980s the British manufacturers of processed cattle food began to incorporate misshapen prions into their product. It might have been because a change in the processes in rendering factories followed a fall in the price of tallow. It might have been because rising numbers of old sheep found their way into the factories thanks to generous lamb subsidies. Whichever was the cause, the wrong-shaped prions got into the system: all it took was one highly infectious animal, riddled with scrapified prions, rendered into cattle cake. No matter that the bones and offal from old cows and sheep were boiled to sterility as they were rendered into protein-rich supplements for dairy cattle.
Scrapified prions can survive boiling.
The chances of giving a cow prion disease would still have been very small, but with hundreds of thousands of cows it was enough.
As soon as the first cases of 'mad-cow disease' went back into the food chain to be made into feed for other cows, the chain reaction had begun. More and more prions came through into the cattle cake, giving larger and larger doses to new calves. The long incubation period meant that doomed animals took five years on average to show symptoms. When the first six cases were recognised as something unusual by the end of 1986, there were already roughly 50,000 doomed animals in Britain, though nobody could possibly have known it. Eventually about 180,000 cattle died of bovine spongiform encephalopathy (BSE) before the disease was almost eradicated in the late 1990s.
Within a year of the first reported case, skilful detective work by government vets had pinned down the source of the problem as contaminated feed. It was the only theory that fitted all the details and it accounted for strange anomalies such as the fact that the 2 8 2 G E N O M E
island of Guernsey had an epidemic long before Jersey: the two islands had two different feed suppliers, one of which used much meat and bonemeal, while the other used little. By July 1988 the Ruminant Feed Ban was law. It is hard to see how experts or ministers could have acted more quickly, except with perfect hindsight. By August 1988, the Southwood committee's recommendation that all BSE-infected cattle be destroyed and not allowed to enter the food chain had been enacted. This was when the first blunder was made: the decision to pay only fifty per cent of each animal's value in compensation, thus providing an incentive to farmers to ignore signs of the disease. But even this mistake may not have been as costly as people assume: when compensation was increased, there was no jump in the number of cases notified.
The Specified Bovine Offals Ban, preventing adult cows' brains from entering the human food chain, came into force a year later, and was only extended to calves in 1990. This might have happened sooner, but, given what was known about the difficulty other species had of catching sheep scrapie except by direct injection of brain into brain, at the time it seemed too cautious. It had proved impossible to infect monkeys with human prion diseases through their food, except by using huge doses: and the jump from cow to person is a much bigger jump than from person to monkey. (It has been estimated that intracerebral injection magnifies the risk 100 million times compared with ingestion.) To say anything other than that beef was 'safe'
to eat at this stage would have been the height of irresponsibility.
As far as scientists were concerned, the risk of cross-species transmission by the oral route was indeed vanishingly small: so small that it would be impossible to achieve a single case in an experiment without hundreds of thousands of experimental animals. But that was the point: the experiment was being conducted with fifty million experimental animals called Britons. In such a large sample, a few cases were inevitable. For the politician, safety was an absolute, not a relative matter. They did not want few human cases; they wanted no human cases. Besides, B S E , like every prion disease before it, was proving alarmingly good at springing surprises. Cats were catching it P O L I T I C S 2 8 3
from the same meat and bonemeal that cattle ate - more than seventy domestic cats, plus three cheetahs, a puma, an ocelot and even a tiger have since died of B S E . But no case of dog B S E has yet appeared. Were people going to be as resistant as dogs or as susceptible as cats?
By 1992, the cattle problem was effectively solved, although the peak of the epidemic was still to come because of the five-year lag between infection and symptoms. Very few cattle born since 1992
have caught or will catch B S E . Yet the human hysteria was only just beginning. It was now that the decisions taken by politicians started to grow steadily more lunatic. Thanks to the offals ban, beef was now safer to eat than at any time in ten years, yet it was only now that people began to boycott it.
