Theses Doctoral

Engineering the Repeats-in-Toxin Domain for Biotechnology Applications

Shur, Oren

Repeat peptide domains are ubiquitous in nature. They consist of tandem repeats of a consensus sequence and are involved in a very wide array of different functions. The predictability of their sequence and, thus, their secondary structure has made them appealing scaffolds to bioengineers. A number of commonly found motifs have been subjected to bioengineering work aimed at consensus design. Such work focuses on identifying a modular repeat unit that can be multiplied as necessary to create a functional interface for applications such as creating novel biomolecular interactions or biomaterials. To be successful, a thorough understanding of the target motif must be obtained. One must understand what amino acids can be targeted for engineering and which are critical for proper folding. In this dissertation we describe our efforts to develop a new scaffold for biotechnology applications based on the beta roll forming repeats-in-toxin (RTX) domain. Unlike existing scaffolds, which tend to be static in structure, the RTX domain is a stimulus responsive domain which only forms a beta roll in the presence of calcium. It is for this reason that we have focused our efforts on this domain. The first three chapters of this dissertation describe our efforts to gain an in depth understanding of the RTX domains folding behavior and amenability to engineering. The final two chapters present two examples of using engineered RTX domains for two different biotechnology applications.
RTX domains are known to only fold when flanked by a protein "cap". In Chapter 2, we study this requirement in detail. A variety of different sized flanking groups are tested on an RTX sequence from the adenylate cyclase toxin (CyaA) of Bordetella pertussis. We begin by testing different truncations of the natural flanking sequence that is found on this RTX domain within the CyaA protein. We show that only the C-terminal flanking sequence is required for calcium-responsive folding. Then, alternative capping groups are tested to study whether or not the native flanking sequence is necessary or any cap is acceptable. We find that the maltose binding protein and certain fluorescent proteins can act as caps that enable folding, and further that these are only required on the C-terminus.
For many biotechnology applications, the ability to tether a peptide on a solid surface is of value. Therefore, in Chapter 3 we investigate if the RTX domain can form a beta roll when immobilized on a solid surface. We use a quartz crystal microbalance (QCM) with RTX peptides immobilized on gold. By testing the C-terminally flanked RTX domain and RTX domains immobilized on either termini, we find that that a solid gold surface can also act as a cap that enables folding. Further, it is found that, unlike with protein caps, tether on either terminus can enable calcium-responsive folding.
Having studied the capping requirements of the RTX domain in detail, Chapter 4 addresses the modularity and ordering of the actual RTX repeats. Both capped and uncapped RTX domains are synthesized consisting of three different lengths. Also, it is noted that there are both "standard" and "non-standard" RTX repeats, so we test the impact of orienting them in different positions relative to the flanking group. It is found that altering the number of RTX repeats still enables calcium-responsive beta roll formation, but with a slightly compromised conformation. On the other hand, we show that altering repeating ordering is catastrophic to beta roll function and only the addition of more repeats can recover some of the native folding behavior.
In Chapter 5, we make our first attempt at engineering novel functionality into the RTX domain. The amino acids that correspond to the beta sheet face are randomized and incorporate into two directed evolution display systems in order to try and select for streptavidin binding beta rolls. We explore both bacterial cell surface display and ribosome display. Using bacterial surface display, we show that the beta roll (both capped and uncapped) can be expressed on the cell surface, but significant problems are encountered with clonal selection of non-beta roll expressing cell such that this approach is abandoned. Using ribosome display, we identify a set of possible streptavidin binding beta rolls. One in particular is shown that may have a low micromolar binding affinity that is calcium-responsive. Finally, structural studies are performed on this clone to assess if it indeed forms a beta roll.
In preparing for the work in Chapter 4, it was discovered that a "consensus" designed RTX sequence undergoes calcium-responsive precipitation. Therefore, chapter 6 presents an effort to develop this into a useful technology for protein bioseparations. We show that three different proteins can be purified using this beta roll tag (BRT) and that the system can be coupled with a specific protease to purify the maltose binding protein. It is demonstrated that by altering the size of these BRTs, precipitation behavior can be modulated.
Overall, in this dissertation we demonstrate the full cycle of developing a new scaffold for bioengineering. We begin by identifying an interesting naturally existing repeat domain. Then, we continued by characterizing, in detail, its folding behavior and tolerance to engineering. Finally, we conclude by showing two different applications that leverage our earlier knowledge to create novel, useful biomolecules for biotechnology applications.


  • thumnail for Shur_columbia_0054D_10880.pdf Shur_columbia_0054D_10880.pdf application/pdf 4.84 MB Download File

More About This Work

Academic Units
Chemical Engineering
Thesis Advisors
Banta, Scott A.
Ph.D., Columbia University
Published Here
April 16, 2014