Antibody regions that fit snugly against the antigen, shown in blue, are more abundant in the antibody with higher affinity (bottom).


Using X-ray crystallographic snapshots of antibodies with increasing affinities for a protein antigen, Yili Li, Roy Mariuzza (University of Maryland Biotechnology Institute, Rockville, MD), and colleagues suggest that protein–protein interactions are optimized by increasing hydrophobic stickiness and improving the fit between proteins.

Protein interactions are optimized by evolutionary changes that enhance the binding energy between the relevant molecules. The immune response offers a unique opportunity to study these changes in a practical time span. During affinity maturation, B cells produce antibodies with increasing affinity for the antigen—a sort of rapid molecular evolution resulting from somatic mutation of the antibody genes. Mariuzza's group examined the structural differences between four antibodies against a lysozyme antigen to determine how the antibodies improved their antigen-binding abilities.

They found that the number of hydrogen bonds and van der Waal contacts, often thought to be the most critical interactions at protein–protein interfaces, did not correlate with improved binding. Instead, hydrophobic interactions were key. As the antibody's ability to bind the antigen improved, an increasing amount of hydrophobic surface was buried at the interface. The alterations also improved shape complementarity, thus filling energetically unfavorable cavities in the interface.

The residue changes that increased hydrophobic interactions and improved complementarity occurred not in the center of the contact interface, but rather at the edges. “At the center, interactions are already optimized by the germ line– encoded antibody,” says Mariuzza. “There's no need to change those through somatic mutation. You must improve the parts that are less than ideal. That's why optimization occurs at the periphery.” Thus, to engineer antibodies with higher affinities to target proteins, researchers should perhaps focus on mutating peripheral contacts. ▪


Li, Y., et al. 2003. Nat. Struct. Biol. 10.1038/nsb930.