Telling the sheep from the cows

When scientists testing for BSE got sheep's brains mixed up with cows', it was a major scandal. But, Steve Connor says, maybe they were right after all

As scientific cock-ups go, this one had true comic potential. For more than four years, scientists were testing sheep brains for evidence of BSE, that is until another team of researchers pointed out that the brains came from cows. It even prompted one wag in a newspaper to publish "a handy guide for ministers and scientists" on how to tell a sheep from a cow – it consisted of a picture of each animal with the captions "this is a sheep" and "this is a cow".

But the sorry, if humorous, saga could have an even more bizarre conclusion. A new set of tests now being planned might end up redeeming the original experiment into whether British sheep have been infected with BSE. Fresh DNA tests on frozen samples of brain tissue used in the original experiment could show that the brains did come from sheep after all.

If this proves to be the case it will create a different sort of headache for government scientists and ministers. It will mean the original experiment's results, suggesting BSE may have infected sheep in the late 1980s and early 1990s, still stand.

The issue centres on a collection of 2,867 sheep brains that were collected between 1990 and 1992 as part of research into the effects of meat-rendering practices on the infectious agent behind scrapie, a sheep disease similar to BSE. The brains were collected in order to investigate how well the scrapie agent survives rendering. Having finished this research, and a similar experiment using cattle brains, the researchers froze the material left over and stored it in a lab freezer at the Neuropathogenesis Unit in Edinburgh, which is part of the UK's Institute for Animal Health (IAH).

In 1996, when the link between BSE and the human brain disease vCJD was established, scientists took a renewed interest in these sheep brains, which had been made into a "pool" of material resembling cold porridge. They thought it could be tested to reveal whether sheep in the national flock have been infected with BSE at a time when the cattle epidemic was at its height.

This was more than idle curiosity. Results of experiments published in 1996 showed that it was relatively easy to transmit BSE to sheep by deliberately feeding them brain material from infected cattle. It was also known that sheep in the late 1980s and early 1990s were given feed that must have been contaminated with BSE. It was therefore possible that some sheep with scrapie might be suffering from BSE, given that the symptoms of both are similar in sheep.

The Government's Spongiform Encephalopathy Advisory Committee pressed the Ministry of Agriculture, Fisheries & Food (Maff) to test the sheep brain pool kept from the 1990-92 experiment. As a result, Maff commissioned scientists at the IAH to carry out the study, a project that began in 1997.

The complicated experiment involved injecting small amounts of the material into genetically distinct strains of laboratory mice which would incubate the brain disease in a predictable way if the BSE infectious agent was present. Normally, using BSE from cattle, the mice show this distinctive incubation pattern straight away, but in this particular experiment the results were equivocal and a second "passage" had to be undertaken – when brain material from one set of mice were injected into a second set.

The results of using this second passage were not simple to interpret but scientists who are familiar with the findings believe they might indicate that BSE was indeed present in the original sheep brain pool.

One important question, however, was whether some sheep whose brains were added to the pool were genuinely infected with BSE or whether BSE from cattle brains had cross-contaminated the sample. This latter scenario was always a possibility given that the sheep brains in the rendering experiment were collected by veterinary scientists who may have used the same surgical slabs and instruments to remove the brains of BSE-infected cattle and scrapie-infected sheep. The rendering experiment was not designed with the same anti-contamination precautions needed for the "strain typing" mouse experiment.

To begin answering the question of contamination, scientists from the IAH carried out an initial test on the brain pool to search for the presence of a unique sheep "signature". Only sheep have the amino acid arginine at position 171 on the prion protein implicated in scrapie and BSE. This diagnostic, called the arginine-171 test, proved positive, indicating that the pool of tissue was, at the very least, substantially sheep in origin.

Professor Chris Bostock, the head of the IAH, says the arginine-171 test has proven to be reliable and he has no reason to doubt the results. "The conclusion that there was 171 arginine in that material looks pretty convincing to me," he says.

However, as we now know, the test was dramatically contradicted by the findings of a more extensive series of DNA tests carried out by the Laboratory of the Government Chemist (LGC), a government-commissioned private institute. These tests indicated that the "sheep" pool was in fact 100 per cent cattle in origin. Helen Parkes, the forensic scientist in charge of the tests, says that the LGC is convinced of the integrity of its methods. "It's the sort of evidence we would go to court with," Dr Parkes says. Indeed, there seems little doubt about the accuracy of the LGC's findings. Dr Parkes was sent triplicate samples of the "sheep brain pool" from Edinburgh and she had analysed each one eight times. They all gave unequivocal results: only bovine tissue was present.

Dr Parkes says that although the frozen samples had thawed out on their 48-hour journey from Edinburgh to the LGC in Teddington, this would have made no difference to the result. The DNA test has proven to work accurately on samples that had been frozen and thawed at least 10 times, she says. "Another cycle of freezing and thawing would have made no difference. We're quite confident of the robustness of these techniques," says Dr Parkes.

There is, nevertheless, one possibility that could still make the original experiment valid and could explain the discrepancy between the LGC's DNA test showing only cattle brains were present and the arginine-171 test of the IAH showing the brain material was largely sheep.

It is possible that the sample labelled "sheep brain pool" which had gone to the LGC in Teddington might not be from the same pool used in the mouse experiment at the IAH in Edinburgh. The error could be down to a mix-up in what was sent to Teddington, and not what was used in the Edinburgh experiment.

An independent audit ordered by the institute was set up to find the batch of pooled brains used in the original experiment. Dr Parkes says that she has been warned to expect to have to carry out more tests on this newly discovered material to see if there is any sheep DNA present. "It could take no more than two to three days to complete," she says.

Meanwhile, Professor Bostock has confirmed that more tests will now take place. "There are several samples that are key at various points in the history of this material," he says. "I would anticipate testing all of those key samples in a completely robust way that satisfies everybody that the result is correct."

If the results are positive then it will rekindle interest in the results of the study to see whether BSE has infected lamb. If they are negative, then it will be the final nail in the coffin of this experiment and we will once again be no nearer to answering the question of whether the sheep flock has ever been infected with BSE in the past, or whether British lamb is still infected.