Winter's way: why Domantis still leads in antibody technology

Engineer Greg Winter spends his days building, folding chains and working with scaffolds. He's also a librarian. But rather than cataloguing hefty manuscripts, he has viruses create each new inventory for him -- an elegant manufacturing process if ever there was one.

Winter does his engineering with antibodies -- creating catalogues of the immune system's most canny defenders -- which can be screened for their ability to latch on to cell surfaces or other molecules.

Antibodies are proteins that bind to specific targets. This specificity can be harnessed to treat a wide variety of disease. Rheumatoid arthritis, for instance, is now treated with antibodies that bind to and neutralise tumour necrosis factor (TNF), a signalling molecule that aggravates inflammation. The new anti-TNF drugs based on Winter's technology include Humira and Remicade, which together garnered more than $4 billion in sales in 2003.

Winter is the author of a technique for mass-producing humanising antibodies which has been used to bring 15 antibody therapeutics on to the market since the late 1970s, a technique he describes as "making antibodies that the human immune system won't make".

The polymerase chain reaction for reproducing DNA is probably biotechnology's most famous technique. But Winter's antibody-producing virus process -- known as phage display -- may win new renown as therapeutic antibodies finally look to be fulfilling their early promise over the next decade.

A random library of genes which code for the variable, active tips of Y-shapes antibodies are inserted into 'phage' -- specialised viruses that infect bacteria. When these modified viruses replicate, they produce enormous, diverse libraries of randomly shaped human antibodies.

To find out which antibodies will bind to a target -- be it a protein expressed on a cancer cell or TNF, an inflammatory molecule -- the target is fixed to the inside a test tube. A library of antibodies can be poured past, and only those which are specific for the target will remain. These can then be isolated and multiplied.

And now Winter believes he has trumped his own invention, with a new process that creates even tinier antibodies which are able to be heated up and frozen and then "snap back into shape" when returned to normal temperature. They're known as 'domain antibodies' -- the sensitive tips of the antibody, one tenth of the size, but just as specific as normal humanised antibodies in their ability to bind to proteins.

Winter explains that during his original phage-display work he noted that antibody single domains could be very potent. Cambridge Antibody Technology -- the company which has commercialised his phage-display technology -- had intended to pursue this new offshoot of the technology, but the single domain antibodies formed a sticky aggregate which was "awful to work with".

But rather than leave the problem in the mire of history, Winter persisted, figuring that the smaller domains would have a huge advantage in be able to fit into the holes and cavities of protein surfaces, binding more tightly to their targets. They are also likely to have better tissue penetration, stay in the body for a longer time, and be more cost-effective, as they have higher yields.

Single-domain antibodies had the added advantage of be able to slip through some legal loopholes, and escape the broad patent coverage Genentech and others had fenced off around monoclonal antibodies.

So after much tweaking, Winter was refined a method for selecting single domain antibodies that preferred not to stick together, which in December 2000 formed the basis for a new company, Cambridge-based Domantis.

Domantis now holds an exclusive license from the UK's Medical Research Council for patents covering creation of antibody libraries (Winter II) including repertoires of single domain antibodies and phage display of antibody fragments.

Last year, Domantis announced it was also exploring an emulsion technology for extracting specific genes encoding proteins with desirable activities from large pools of gene sequences. Although a few years behind phage display, this technology does not use bacteria or cells and has the potential to be automated.

Australian connection

In the hinterground of this rather English story lurks an Australian company, Peptech -- very much the antipodean cousin to both Cambridge Antibody Technology and now to Domantis. Peptech founder Geoff Grigg worked with Winter in the 1980s -- "We managed to fit him into a fume cupboard," says Winter wryly.

By the time Grigg heard about the impending birth of Cambridge Antibody Technologies, he had founded Peptech, and was in the position to provide some of the original seed funding, taking a 40 per cent stake.

The relationship lapsed somewhat, although Peptech continues to earn undisclosed royalties from monoclonal antibody drugs such as Humira. History repeated itself when Peptech bought into the nascent Domantis. It has recently upped its investment to 36 per cent -- a vote of confidence in Domantis' management and its intellectual property.

Domantis also developed Peptech's lead drug, a single domain antibody molecule to treat TNF-mediated inflammatory disorders. Peptech announced last year that the new antibody had demonstrated its promise as a second-generation therapeutic in pre-clinical trials in an animal model of arthritis, proving significantly more potent than other anti-TNF biologics on the market.

Under an agreement signed when it acquired its original stake in Domantis, Peptech has options to develop two more single domain antibody products for therapeutic targets of its own choice.

And having fathered two unique technologies, Winter is still eager to tinker with his antibodies. He speaks warmly about the possibilities for bi-specific antibodies, which would be able to target two antigens, one on each arm of the antibody Y. Normally, antibodies have two identical arms. Perhaps these dual-purpose antibodies might be used to localise the anti-TNF function of monoclonal antibodies -- so useful for in suppressing inflammation -- to a particular region of the body, such as the joints. However, TNF also has the beneficial ability of being able to suppress tumours, and one of the unwanted by-products of anti-TNF therapy -- risking an outbreak of tumours -- would thus be avoided.

Requiring two targets could also make antibodies much more specific for cancer cells, sparing healthy cells.

For now, Domantis has 12 therapeutic programs underway -- aimed at developing domain antibodies drugs for asthma, Crohn's disease, infection, cancer and cardiac disorders.

"We intend to take one or two into humans by 2007," says Winter. And with his track record, you have to believe he will.