In March 2011, the US Food and Drug Administration approved a new drug for the treatment of systemic lupus erythymatosus. Developed by Human Genome Sciences and GlaxoSmithKline, Benlysta targets a protein called BLyS and is, as Nature News pointed out, the first new drug for Lupus “in half a century.”

Benlysta's generic name, belimumab, bears the stem –mab, identifying it as a monoclonal antibody similar to Herceptin (trastuzumab), Rituxan (rituximab), and Remicade (infliximab). But Benlysta is different from those other therapeutics: Its variable region was created not by immunizing a laboratory animal, but in a test tube.

Benlysta was born via a process called phage display. Antibody phage display and related in vitro technologies such as yeast display and ribosome display use vast, more or less random libraries of immunoglobulin variable regions to identify molecular binders to specific antigens. Developed in 1990 and 1991 by groups at the MRC Center for Protein Engineering in Cambridge, UK, The Scripps Research Institute in La Jolla, Calif., and in Heidelberg, Germany, the technique has for two decades been used mostly to drive therapeutics development. Now, though, loosening intellectual property entanglements are freeing the antibody development community to embrace the technology as never before. European and American funding agencies have established projects to build high-throughput pipelines for the creation of in vitro antibodies, and plans are in the works to develop antibodies for the entire human proteome.

As Stefan Dübel of the Technische Universität Braunschweig, a member of the Heidelberg team, succinctly puts it, “They are now absolutely ready for prime time.”

Phage Display 101

Phage display is a technique so simple in concept, it's genius. According to John McCafferty, Director of Research at the University of Cambridge and lead author on the first phage display article from the Cambridge team in 1990, the general idea is to establish a screening system in which each antibody in the screening library ferries around its own genetic code.

“That's really what display technologies are all about,” he says, “this joining a gene and the product of that gene in a way where you can use the binding properties [of the antibody] to help you get to the gene.”

The inspiration came from a paper by George Smith at the University of Missouri, Columbia. At the time, McCafferty and his colleagues in Greg Winter's lab were interested in human antibody therapeutics, but the process was long — generating a mouse monoclonal, which takes months in itself, was bad enough, but then “humanizing” it piece by laborious piece was painfully slow. In 1988, Smith showed he could modify filamentous bacterial viruses, or bacteriophage, with a piece of the beta-galactosidase gene, which would then be displayed on the virus coat as a fusion with a minor coat protein. He used antibodies to beta-galactosidase in a “biopanning” experiment to recover beta-galactosidase-expressing phage, and speculated the approach could be used more generally to clone genes for which the protein was on hand but not the DNA sequence.

“We looked at that and thought, wouldn't it be cool if you could turn that all the way around, [and] have the antibody on the phage and have the antibody recognizing targets?” recalls McCafferty.

So, he did. First, McCafferty demonstrated a single lysozyme-targeting antibody could be expressed on the phage surface. But it wasn't any standard immunoglobulin like those found in serum; it was a kind of ‘Franken-body’, a chimeric molecule in which the variable light chain and variable heavy chain portions of the standard immunoglobulin were tied together using an artificial linker. These constructs, called “single-chain Fvs,” or scFvs, are easier to manipulate and express in bacteria than intact immunoglobulins, and work just as well: In the presence of a million-fold excess of non-specific phage, the team was able to pull out viruses specific for lysozyme. (An alternative antibody format for display technologies is the Fab fragment, one arm of an intact Y-shaped immunoglobulin molecule.)

The following year, James Marks, with McCafferty, Winter, and colleagues, took the study to its logical conclusion, replacing the anti-lysozyme fragments with a random library of human antibody pieces from which binders to any antigen could, in theory, be pulled, and the modern phage display process was born.