By David N. Leff

Are the protein engineers who whittle down monoclonal antibodies to leaner, meaner, high-performance fragments painting themselves into corners?

In one corner: Many of the truncated antibodies under development are missing the Fc protein region that adorns complete, natural antibodies. This is a real loss, because the Fc portion is what sets off the immune system's effector functions its two prime weapons against infection -- namely, the complement cascade and phagocytes/monocytes.

Complement terminates cells by punching holes in their walls; phagocytes by engulfing and digesting them, and clearing their corpses from the circulation.

A second corner is the absence of glycosylation in engineered fragments.

Bacterial fermentation is an increasingly important method for generating antibodies, but bacterial expression requires that the Fc region be glycosylated before it can interact with its complement or cellular receptors.

The Fc receptor triggers processes such as the phagocytosis of target cells. "Say you have a virus or bacterium that's been decorated by antibodies," explained molecular biologist Greg Winter. "Its Fc portion is recognized by Fc receptors on the phagocytes. They bind, engulf and digest those targets."

A third corner, of vital concern in immunotherapy, is keeping antibodies in the bloodstream on the job. "One of the problems in dealing with fragments," Winter told BioWorld Today, "is that these tend to get lost very quickly from the serum compared to complete antibodies, which survive much longer."

Winter is senior author of two back-to-back articles in the July issue of Nature Biotechnology, which report his group's novel strategy for dealing with these fragmentary shortcomings. One is titled: "Complement recruitment using bispecific diabodies;" the other, "Retargeting serum immunoglobulin with bispecific diabodies."

He heads the division of protein and nucleic acid chemistry at the Medical Research Council's Laboratory of Molecular Biology, University of Cambridge.

Diabodies, Anyone?

Here's how Winter explains diabodies, which he invented in 1992:

"What we've done with the diabodies is to make a scaled-down version of the antibody, which is also capable of recruiting either the Fc receptors, or the complement directly. No one, to our knowledge, has demonstrated the direct recruitment of soluble complement, as we have just reported in Nature Biotechnology."

He continued: "We have used diabodies to recruit straight serum antibodies to knock out the target tumor or infectious cell, utilizing effector. One of its two arms is directed against our target cell, the other against a quite different antigenic epitope on serum immunoglobulin.

"What we can do is take those antibodies that are against one specificity and by making an adapter molecule, a diabody, we can give them a second specificity, directed against a quite different target of interest, whether tumor cells, bacteria, viruses or whatever. "

Winter cited a further advantage of diabodies: "Because we're binding to antibody or the complement cascade, both soluble components in the serum, this tends to hold on to the diabodies, and serum half-life is thereby greatly extended. Obviously," he observed, "they have much more therapeutic effectiveness if they hang around that much longer."

Diabodies, Winter pointed out, "are only a third the size of native antibodies. Therefore, they should penetrate tissues more effectively. One of the problems with native antibodies is that they are so huge, it's difficult for them to pass into tissues, through blood vessels. Therefore, the smaller the fragments are, the better.

Antibodies, which are two-pronged, Y-shaped molecules, elicit a complement cascade attack by landing with both arms on the surface of a cell marked for destruction. "What we're doing," Winter explained, "is putting an additional kind of stick-on there -- a diabody -- so you get a tripod. And that added arm has a new specificity. One arm of the diabody is directed against a portion of the antibody's light-chain constant site, which calls in the effector, while the other arm provides the novel binding site.

"The biotechnologists," he observed, "are interested in recombinant antibodies being produced in bacteria. But you can't produce complete antibodies, as we've seen, because these have to be glycosylated to recruit effectors, and the yields from bacteria are quite terrible.

"So another interest in having fragments rather than complete antibodies is that there's a lot of new technology now for cloning fragments in bacteria in large quantity and with much lower cost than having to make complete monoclonals in eukaryotic cells."

As for the therapeutic potential of diabodies, Winter said, "We can only speculate. We demonstrated the principle here," he added, "and the next step, I think, is for someone else to take this on therapeutically.

"In fact," he went on, "the patents protecting this diabody technology have been assigned by the British Medical Research Council to Cambridge Antibody Technology Group plc [CAT], of Royston, Cambridgeshire. I was one of the founders of that company."

Exciting Commercial Potential, But No Trials Yet

CAT floated on the London Stock Exchange March 25, 1997, its CEO, David Chiswell, told BioWorld Today. (See BioWorld International, June 4, 1997, p. 3.)

"We think diabodies are very interesting molecules, and with exciting commercial potential," he observed, "but at this point we have no announced trials with diabodies. Our stated goal," Chiswell went on, "is to build up the company so that we can develop three to four new clinical programs every year from the year 2000.

"We haven't been specific about the new programs we would start to do in '99 and 2000." he pointed out. "From then on, we aim to do three to four new programs per year. With any luck, we would hope that some of those would be diabody-formatted."

Winter said that preclinical in vivo studies of diabodies in mice with implanted human tumors "is not work that we ourselves would tend to do here. If we were to go down that route, we would want to identify suitable targets and suitable collaborators. And that's a commercial matter, best done in collaboration with CAT." *