Fusion is the concept of the hour, supposedly the key to everything from better cuisine to clean energy. Fusion proteins certainly have shown that they have potential in medicine, since humanized antibodies consist of mouse and human proteins that have been fused at the genetic level.

In the May 2, 2006, issue of the Proceedings of the National Academy of Sciences, researchers from Immunomedics Inc., its subsidiary, IBC Pharmaceuticals Inc., and the Garden State Cancer Center in Belleville, N.J., reported on a new method that can be used not just with antibodies, but also with any number of proteins and other biological molecules by using two naturally occurring alpha-helices.

Current fusion technologies are either molecular biological, where two proteins are combined at the genetic level and produced by cell cultures, or by old-fashioned chemical conjugation.

However, "recombination is often limited by productivity levels and the level of complexity that you can build into the molecule," lead author Edmund Rossi told BioWorld Today.

Chemical conjugation, on the other hand, typically modifies amino acids - either lysines or cysteines - that are part of the molecule that is to be conjugated, which leads to its own set of problems. "It's very difficult to get a homogenous product, because the change is typically not site-specific," Rossi said. If a protein has several of the amino acids in question, one, several or each of those amino acids may end up tethered to its fusion counterpart. In addition, because the amino acids have day jobs within the protein, chemical conjugation can lead to inactivation, defeating whatever the purpose was of conjugating in the first place.

In contrast, the method described in PNAS adds the cysteines that link the helices to their targets, and orients them in a way that limits their interaction to those desired by the researchers.

The scientist used two linker proteins, the dimerization and docking domain of protein kinase A (DDD) and the anchoring domain of A-kinase anchoring proteins (AD). The final construct will contain two copies of whichever protein fragment is linked to DDD and one copy of the fragment linked to AD, because DDD always forms dimers, with each dimer binding to one copy of the AD-linked molecule.

In the end, the method gives "one final product consisting of two components with three functional groups," said Chau Cheng, associate director of investor relations and business analysis at Immunomedics.

In the PNAS paper, the researchers validated the technology in vivo by using it for pretargeting cancer therapeutics. When they injected a fusion antibody with three binding sites - two for the tumor and one for a radiotracer - followed by the radiotracer itself once the first antibody had bound to the tumor, 30 percent of the tracer was bound in the tumor within an hour after injection, and blood-to-tumor ratios of up to nearly 400 a day after injection suggest that the technology could be used to send toxins into tumors with reduced toxicity to other tissues.

Immunomedics and IBC plan to both out-license the technology and develop their own therapeutics based on it. Other companies are working to expand the possibilities of fusion. Syntonix Pharmaceuticals' Synfusion technology has nabbed three major deals for its antibody fusion technology within the past 13 months.

Its technology links the Fc region of an antibody to a drug, allowing the drug to take advantage of a biological pathway to create longer-acting proteins or peptides that can be injected less frequently. (See BioWorld Today, April 1, 2005; Oct 24, 2005; and Jan. 24, 2006.)