The development of systems to produce novel recognition functions using nonan-tibody scaffolds has been driven by several motivations. First, the physicochemi-cal properties of natural antibodies are not ideal in many applications in research, biotechnology, and therapy. Immunoglobulins are large molecules (~150 kDa) made of two copies each of the heavy and light chains. Their stability is dependent on intradomain and interchain disulfide bonds and thus they do not fold under reducing conditions. Their production is cumbersome and expensive. The heterodi-meric structure makes reformatting for downstream applications difficult. Smaller fragments of immunoglobulins (e.g., Fab and single-chain Fv) have been used for antibody-engineering applications, but they inherit many of these limitations (Worn and Pluckthun 2001). The hope is that these limitations can be overcome by switching to synthetic binders built on a simpler scaffold. Second, from a basic science point of view, the exercise of engineering synthetic interfaces critically tests and enhances our understanding of the principles governing molecular recognition and explores the capacity of protein engineering technology. Third, the intellectual property landscape associated with recombinant antibody engineering is so complex that it has become practically impossible to sort out the situation. This cause is not intellectually stimulating, but nevertheless critically important in the industry. From a commercial point of view, an "alternative" scaffold system with little prior intellectual property would be highly attractive.
Although the concept of creating a synthetic recognition interface on a scaffold is simple, it is technically daunting, because one essentially needs to recapitulate a directed evolution system that rivals that used by the adaptive immunity. Such a system minimally consists of a "library" containing large sequence diversity and a method to isolate members with a desired property. Fundamental technological breakthroughs that have established the field of synthetic interface engineering are collectively known as molecular display technologies. These include phage display, mRNA display, yeast surface display, and yeast two-hybrid, which all establish unambiguous linkage between protein phenotype and genotype. Such phenotype-genotype linkage is critically important for library selection and directed evolution, because one ultimately needs to isolate (i.e., clone) and determine the identity of proteins that have been selected. Equally important, they enable the generation of large combinatorial libraries whose size rivals or exceeds natural immune diversity. Molecular display technologies and library construction methods are described in detail in Chapters 1-4.
Was this article helpful?