Advances In Display Technologies

An important aspect of successful interface design and engineering is the ability to generate and test a large number of amino-acid sequence diversities. Thus, molecular display technologies are an indispensable tool in directed evolution-oriented interface engineering. Phage display (Chapter 1) has been the most commonly used display technology in this field. However, other methods such as yeast display (Chapter 2), and ribosome display and mRNA display (Chapter 3) have been successfully used. In general, nonantibody scaffolds developed for synthetic binders are small, stable, and devoid of disulfide bonds, and thus they are compatible with a broader range of display technologies than are antibody fragments. For example, FN3-based binders have been generated using virtually all display technologies available (Koide et al. 1998; Koide et al. 2002; Xu et al. 2002; Koide et al. 2007a; Lipovsek et al. 2007; Garcia-Ibilcieta et al. 2008).

A larger starting library can sample a larger sequence space and should in principle provide a greater chance of producing a binder if the same selection or screening procedures are used. In practice, however, most experiments are performed with the largest library size that an investigator can produce with a particular display technology, and it is difficult to determine whether success or failure of an attempt can be attributed solely to the library size. For example, although it is difficult to produce a library greater than 108 using yeast surface display (a level that is potentially five orders of magnitude smaller than those constructed with ribosome display or mRNA display), highly functional interfaces have been engineered using this technique (Lipovsek et al. 2007). The quality of a library is equally important for successful generation of functional molecules. A larger library containing a large fraction of "junk" (e.g., stop codons and frameshifts) may be outperformed by a much smaller one in which most library members are functional molecules. It can also be useful to combine different display methods, for example, phage display and yeast display, to exploit their complementary strengths (Koide et al. 2007a; Koide et al. 2007b).

In the conventional filamentous phage display systems, proteins are fused to phage coat proteins that are secreted into the periplasm through the SecB-mediated posttranslational secretion pathway. In this secretion mechanism, fully translated proteins are threaded through the membrane in an unfolded state (Sidhu and Koide 2007). While this mechanism is suitable for proteins that are marginally stable in the bacterial cytoplasm, such as antibody fragments, it presents a major bottleneck for the display of stable proteins that fold rapidly in the cytoplasm (O'Neil et al. 1995). Because nonantibody scaffolds often fall in the latter category, this bottleneck is potentially a serious problem.

This secretion bottleneck has been overcome by use of the cotranslational secretion pathway mediated by the signal recognition particle (SRP) (Steiner et al. 2006). Use of the SRP-dependent pathway increased display levels of highly stable proteins, such as DARPins, as much as 700-fold relative to the Sec pathway. The SRP-dependent phage display system was also effective in increasing the display level of another highly stable scaffold, FN3 (Koide et al. 2007a). Thus, the use of SRP-dependent systems considerably broadens the applicability of phage display. An interesting possibility is that a highly promising scaffold may fail because of a low level of phage display caused by its high stability and rapid folding. This introduces a caveat to high scaffold stability, which, as indicated earlier, is an otherwise favorable property of a molecular scaffold.

Compared with phage display, relatively small numbers of scaffold systems have been tested with other display technologies such as yeast surface display and mRNA/ ribosome display. Thus, it is difficult to determine pros and cons of these display technologies in engineering synthetic binders based on nonantibody scaffolds. Potential problems with yeast surface display include smaller library sizes and non-natural glycosylation. In addition, although early work suggested that yeast surface display could be used to improve protein stability, recent reports demonstrated that this capacity is rather limited, particularly when applied to a small protein (Park et al. 2006; Dutta et al. 2008). Potential problems with mRNA and ribosome display include the difficulty to produce protein homo- and hetero-oligomers. Together, it is extremely important to consider the biology and potential constraints of a molecular display technology when developing a new scaffold.

0 0

Post a comment