Proteus pins hopes on gene discovery

Paul Rodgers explains why large drug firms suddenly want to talk to a small biotech company

LONG-SUFFERING investors in Proteus, a biotechnology company, may be sceptical, especially when they see its financial results on Thursday. But Jurek Sikorski, the new chief executive, is confident the company is about to turn the corner, thanks to a new technology he believes is capable of targeting the genes behind diseases ranging from rheumatoid arthritis to cancer and Aids.

Proteus, like most biotech companies, has had a turbulent life. Floated in 1990 at 84p, its shares soared to about 500p two years ago, then collapsed after UBS quit as joint broker when it failed to meet projections in the original prospectus. The stock is now languishing back at 88p, and the markets are braced for another annual loss, probably around pounds 7m, to be announced on Thursday.

But the company, based in Macclesfield, Cheshire, is now in serious talks through Progene, a joint venture with America's Genelabs Technologies, with eight international drug companies about applying its computer design techniques to formulate molecules that can block individual strings of DNA. If even a few of them sign development agreements, Proteus will have a healthy stream of revenues for the first time.

Promises of deals to come are a dime a dozen in the biotech sector, but Proteus's claim is more than just hot air. The head of cancer research at one US drugs giant called the advance the best thing he had seen in 10 years. "They came to us two years ago and, in an appropriately arrogant fashion, we kicked them out, saying that what they proposed couldn't be done. They came back 18 months later with the evidence that they'd succeeded."

Deals between other small biotechnology firms and large pharmaceutical companies typically involve a one-off cheque at the start of a development programme, and additional payments as set milestones are passed. For shareholders, the big payback will only come after a drug gets regulatory approval for use on human patients and fat royalties start to flow.

Proteus had originally aimed to sell the computer software that biochemist Dr Barry Robson began developing at the University of Manchester to pharmaceutical companies. But when they balked at the price tag, it decided to apply the system itself. Some 30 projects were spawned, fragmenting the research effort. When Mr Sikorski was brought in nine months ago, he pared the number of projects down to a dozen. Among them was Progene.

Many diseases have links to DNA, the spiralling ladder-like molecule in the nucleus of each cell that governs its biochemistry. Different sections of the molecule - the genes - are coded to represent proteins, the building blocks of life. Most of the genes are necessary for the cell's well-being but a few, such as cancer-causing oncogenes, are not. The code is written with four base-pairs, each one representing a slightly different-shaped rung on the DNA ladder.

Each gene is switched on and off by molecules called transcription factors, which bind to the DNA in front of the protein-coding sections, like grotesquely shaped men with their arms and legs wrapped through and around the rungs of a ladder. Some naturally occurring drugs also bind to these sites, blocking the transcription factors and thus the copying process.

These molecules are small, however, usually covering just four of the DNA molecule's rungs, so they have a wide choice of possible binding sites. A drug that was 12 to 16 base-pairs long is so complicated it would fit snugly only when it found the target segment. It would therefore be more effective, require a lower dose, and be less likely to cause side-effects.

Progene set about building such molecules by testing known DNA binding drugs with a patented Genelabs process called Merlin to see which combinations of base-pairs they stuck to most firmly. Each four-rung drug was tried against every possible combination of four base-pairs, 256 in all. The result was a catalogue of about 30 building blocks.

Proteus then used its custom computer software, Prometheus, to combine the building blocks into 12 base-pair molecules that match specific strings of DNA. The automated design process usually comes up with several candidates, which are then built and sent to Genelabs for further testing to ensure the new drugs do block transcription factors.

If the large drug companies sign on the dotted line, their role will be to supply Progene with DNA strings that they have linked to specific diseases. The drugs will still have to show that they can pass through the blood stream without breaking down, and that they can sneak through the cellular and nuclear membranes to reach the DNA. Finally they will have to go through tests to prove that they are safe.