The Power of Display
In vitro strategies are not exclusively driven by phage display; other options include ribosome display, in which an antibody-encoding mRNA is tethered via a ribosome to the antibody it encodes, and yeast display, which expresses antibody libraries on the surface of yeast.
Typically, yeast libraries are less diverse than phage libraries, a consequence of the number of transformants the organism can produce. Yet, according to Eric Shusta, associate professor of chemical and biological engineering at the University of Wisconsin, who uses yeast display, the strategy does have its advantages.
Yeast display experiments are scored via flow cytometry, by mixing the yeast population with fluorescently labeled antigens. As a result, the approach quantitatively scores every clone, whereas phage display screens often rely on qualitative or semi-quantitative validation of smaller numbers of clones, typically by ELISA. In addition, because a flow cytometer can be tuned to select only, say, the top 5% of binders in a population, experiments can be designed to identify not just binders in general, but the highest quality binders — those that can bind antigen at low concentrations. As a result, some researchers have begun blending the two techniques, using the larger size phage libraries for initial panning rounds, then yeast display for more efficient screens.
“What's really nice is the ability to do quantitative discrimination of antibody binding parameters directly in line with the screen,” says Shusta, who has developed an approach in his own lab to probe his yeast display libraries with detergent-solubilized cell extracts in order to find membrane protein binders, such as components of the blood-brain barrier that typically challenge phage-based strategies.
Of course, whichever strategy is used, all in vitro display methods offer advantages over traditional antibody methods. Polyclonal antibody preparations are easy enough to make, but the resulting sera are ill-defined, and more significantly, finite: Once that antibody preparation is gone, it's gone; the only way to make more is to inject another animal, which may or may not provide a serum of equivalent quality.
Monoclonal antibodies, generated through hybridoma technolology, are renewable. But the process is time consuming and laborious, taking months to complete. In contrast, in vitro screening can be done in a week or so, whittling libraries containing as many as 100 billion virus particles down to a handful of candidate binders.
Yet in vitro display technologies aren't simply more convenient than hybridoma technology — they can yield antibodies that cannot be generated any other way, such as antibodies to proteins that are exceptionally well conserved between mouse and man.
In vitro display also enables selection strategies that would be difficult to pull off in immunized animals. For instance, according to Karen Colwill, a staff scientist at the Samuel Lunenfeld Research Institute, researchers could use phage display to generate antibodies specific for a protein in its complexed state, but not its uncomplexed one, a screen that would be nearly impossible via immunization because once the complex is injected into the animal, it would be processed into peptides and fall apart.
Another advantage: the coupling of antibody and gene simplifies downstream genetic manipulations, such as applying protein tags, reassembling intact immunoglobulin genes, and “affinity maturation” — shuffling the genetic deck to boost an antibody's binding kinetics. In one such study, Dübel and his team tuned an scFv antibody to MUC1, a protein overexpressed in breast and ovarian tumors, by amplifying the gene using “error-prone PCR.” By sifting the resulting libraries through additional rounds of selection, the researchers created an antibody whose affinity was more than 500-times stronger than what they started with.
Finally, in this age of low-cost synthetic genes, in vitro approaches offer one additional advantage: the ability to distribute sequences rather than physical antibody preparations. “We can send it out in digital form,” says Susanne Gräslund of the Structural Genomics Consortium. “People could synthesize the antibodies themselves in their own labs.”