Milking them for all they're worth

Genetically engineered animals are yielding human proteins to combat hereditary disease - and to make a fortune for the breeders. Stephen Day reports

The most closely watched sheep in the world are probably a flock grazing at a location near Edinburgh. Each identified by its own implanted microchip, these sheep have the biotech equivalent of the golden fleece. They carry the human gene for a protein called alpha-1-antitrypsin (AAT) and secrete large amounts of human AAT into their milk. With the help of an pounds 8m milking parlour and neighbouring extraction plant, PPL Therapeutics, the sheep's owner, is now preparing to test an AAT aerosol to combat the symptoms of cystic fibrosis.

Showing that a good pun sticks, PPL's route into the pharmaceuticals market is generally called "pharming". The company is far from the only "pharmer" out there. Small flocks and herds of genetically engineered cows, sheep, goats, pigs and even rabbits are being milked for human proteins at various secure farms around the world.

The lure of "pharming" is that genetically engineered animals can produce complex human proteins that are either impossible or impractical to make by any other method. Some simple compounds such as insulin are now made by vats of genetically engineered bacteria, yeast or animal cells.

However, these systems cannot produce the most complex human proteins efficiently. In contrast, farm animals secrete complex proteins into their milk naturally. Companies like PPL have hijacked this process by hooking human genes on to appropriate controlling regions of DNA. Injecting the human gene into the DNA of a fertilised egg, and placing the egg in a surrogate mother, can then produce a genetically engineered animal that secretes a human protein into its milk.

It's a painstaking business. The injected gene may not integrate into the egg's DNA and, even if it does, the animal may not produce much of the human protein. But once researchers create even one high-yielding animal, conventional breeding opens the way to an unlimited supply - the inserted human genes, being incorporated into the sheep's genome, can pass from generation to generation like normal genes.

Some of the sheep grazing near Edinburgh are the great, great granddaughters of PPL's original genetically engineered sheep. The AAT-producing flock now numbers around 200 and is giving enough AAT for clinical trials to begin, probably around the end of this year.

According to PPL's managing director, Ron James, the AAT protein cannot cure cystic fibrosis, but it may help prevent one of the disease's most damaging symptoms - the destruction of the lung lining. The natural function of AAT is to inhibit the activity of an enzyme called elastase that the body uses to destroy dead or damaged tissue. "The role of elastase is to chew up bits of tissue as part of a process called tissue remodelling," says James. "This is particularly important in the lungs." In people with cystic fibrosis, however, frequent lung infections lead to an influx of disease-fighting white blood cells which release extra elastase into the lungs.

Normally, the extra elastase is useful because it removes disease-damaged tissue. But after repeated infections it causes a breakdown in the lining of the lung that is eventually fatal. PPL's trial will see whether patients with cystic fibrosis can rein in the effects of the excess elastase by spraying AAT directly into their lungs. According to James, the initial trial will be very small, involving tens rather than hundreds of patients. If the early results are positive, the company may move to larger trials in the middle of next year.

However successful the trial may be, PPL's sheep have a secure future in providing AAT to treat another lung disease, hereditary emphysema - a fatal disorder causing lung damage similar to that resulting from cystic fibrosis. It stems from a genetic inability to produce sufficient AAT and often remains unnoticed until middle age when lung damage results in shortness of breath. In cystic fibrosis, the damage is caused by too much elastase in the lungs. In hereditary emphysema, the elastase level is normal, but there is not enough AAT. Doctors can sometimes halt lung degeneration in emphysema sufferers by injecting AAT extracted from donated blood, but currently this treatment is available only to patients in the US and Germany: there is just not enough AAT to go around. To treat just one patient for one year requires all the AAT from about 100 litres of blood.

In contrast, just a couple of thousand genetically-engineered sheep would be sufficient to meet the world demand.

This ability to produce large amounts of otherwise very scarce proteins explains why drug-producing livestock roam in other places besides the Edinburgh area. PPL's sister company in Virginia has genetically engineered pigs to secrete an anti-blood-clotting agent called Protein C into their milk. The company is also working on genetically-engineered rabbits and cows.

Another company, Genzyme Transgenics in Massachusetts, breeds genetically engineered goats whose milk holds a different clot-digesting drug, called tissue-plasminogen activator. Finally, Gene Pharming Europe, in Leiden in the Netherlands, is breeding cows that produce the human version of lactoferrin, a protein that captures free iron atoms and binds them in its structure. Lactoferrin is doubly useful because although our bodies can use iron that is bound to the protein, many bacteria cannot. So on the one hand, lactoferrin could help to treat anaemia by delivering iron to patients, while on the other, the protein could inhibit bacterial infections by soaking up iron and starving bacteria of the metal.

Sheep, pigs, rabbits, cows, goats - why are the pharmers using so many different breeds of animal? Part of the reason is the trade-off between speed and size. It takes less time to breed genetically engineered rabbits than genetically engineered cows, but the cows will produce the most milk in the long run. In fact, genetically engineered cows' milk may capture the biggest market of all - baby formula. The major proteins in cows' milk differ significantly from those in human milk. This makes cows' milk less suitable for human babies. It may also contribute to allergies. Pharming could solve the problem.

"I think it's quite feasible that you could make milk that contained human versions of the three, four, or five major proteins," says James. However, completely "human" milk from cows is not on the agenda. According to James, the best approach is to identify the key proteins and replace those. "There are a host of other minor proteins. I don't see any way in the near future that you could get a complete replacement," he says. "Nor do I think it would be necessary."

Despite the medical logic of "pharming", it is still conceivable that the public will reject the approach. To strengthen the industry's hand, PPL points out that there appear to be no effects of genetically engineering their sheep beyond the production of AAT in the animals' milk. Furthermore, lambs drinking the AAT-rich milk grow normally. Since each sheep has the potential to produce about pounds 65,000-worth of AAT per year, PPL has every reason to take good care of them. The genetically-engineered sheep come from scrapie-free New Zealand stock and live on a strictly vegetarian diet.

More cynically, companies like PPL could defend themselves by pointing to more tempting targets for potential opponents to focus on. A few years ago, researchers in Australia produced fast-growing "superpigs" by giving them extra copies of a growth hormone gene. But there were no medical benefits - especially not for the pigs. Unlike PPL's sheep, the only benefit superpigs offer is the chance to produce more pork more quickly.