• Libraries of synthetic antibodies
A major advance in oncology over the last decade has been the emergence of monoclonal antibodies as effective therapeutics. Our lab has been involved in developing the latest frontier in antibody therapeutics: synthetic antibody libraries with man-made antigen-binding sites. We have optimized library methods to allow facile construction of libraries two orders of magnitude larger than was previously thought possible (Sidhu et al, 2004; Lee et al, 2004; Fellouse et al, 2007). Libraries of this type offer numerous advantages over hybridoma technology, and we have developed highly functional repertoires that surpass the natural immune system in terms of diversity and functionality. Using shotgun combinatorial mutagenesis and structural analysis, we have explored the contribution to antigen binding of specific amino acid residues within the antigen-recognizing region of synthetic antibodies, and we have demonstrated that some of these residues had a greater impact on affinity and specificity of binding (Vajdos et al, 2002; Pal et al, 2006; Birtalan et al, 2010, Persson et al, 2012). The identity of residues also plays an important role, with tyrosine being particularly significant, as well as serine and glycine (Fellouse et al, 2004; Fellouse, et al, 2005; Fellouse et al. 2006; Birtalan et al, 2008; Gilbreth et al, 2008; Fisher et al, 2010). The findings are guiding us in developing new and improved synthetic antibody libraries, allowing the selection of antibodies with greater affinities to antigens.
• Alternative antibody frameworks
Synthetic antibody technology gives the freedom to explore other frameworks for their ability to bind antigens. The full-length antibody framework consists of a constant region (Fc) and a variable region (Fab). Our libraries have been constructed mainly on the Fab framework, but Fabs are generally reformatted as a full-length IgG molecules for use as reagents and therapeutic agents. The IgG format presents some advantages, such as better affinity due to its bivalent nature, and high serum half-life due to receptor-mediated recycling. However, other formats are currently being studied as alternatives to IgG. These include smaller fragments, such as single chain fragments of the variable domains (scFvs) and autonomous variable domains (VH) (Bond et al, 2003; Bond et al, 2005; Barthelemy et al, 2008; Tonikian & Sidhu, 2012; Ma et al, 2013). In addition, we are exploring augmentation of IgG molecules with additional binding domains to generate bispecific antibodies.
• Generation of synthetic antibodies by high throughput phage display selections
To produce synthetic antibodies, our antibody libraries are displayed on phage and screened against a desired target antigen in an in vitro setting (Sidhu & Fellouse, 2006; Geyer et al, 2012; Adams et al, 2013; Adams & Sidhu, 2014). Because the DNA sequence encoding each particular antibody is packaged within the phage, it can be easily manipulated. Thus, this technology allows for rapid affinity maturation and specificity optimization of the selected antibodies (Li et al, 2009). Moreover, synthetic antibodies are built on optimized human frameworks that minimize the risk of immunogenicity, making them ideal for therapeutic applications. Over the last 15 years, our lab has made significant improvements to phage display-based combinatorial protein technology (Sidhu et al, 2000; Fuh & Sidhu, 2000; Roth et al, 2002; Weiss et al, 2003; Lee et al, 2004; Held & Sidhu, 2004; Sidhu & Koide, 2007). This technology can easily be adapted to a high throughput pipeline, which has allowed us to augment the diversity of our synthetic protein libraries in a systematic manner. By combining high throughput phage display screening with high throughput sequencing, we have now created a robust combinatorial system with greatly improved performance and utility, not only for synthetic antibodies but also for other protein frameworks (Fellouse et al, 2007; Ernst et al, 2010). Most recently, in collaboration with Dr. Jason Moffat’s lab, we have created a transformative new method, termed “CellectSeq”, which allows direct selection of antibodies against antigens displayed on the surface of cells, circumventing the need for antigen purification and addressing the difficulty of producing antibodies against transmembrane proteins, which represent 70% of current drug targets. Our new method permits the rapid development of synthetic antibodies against any protein expressed in its native form on the cell surface.
This pipeline has been scaled up to form the Toronto Recombinant Antibody Centre (TRAC), a state-of-the-art platform focused on the high throughput generation of a comprehensive repertoire of high affinity therapeutic-grade synthetic antibodies. Using high throughput phage display technology, we have, in a short time frame, produced thousands of high-affinity antibodies with dissociation constants in the single-digit nanomolar range, against a wide variety of protein antigens.
• Modulation of cell signaling with synthetic antibodies, and antibodies as potential therapeutics and reagents for biological research
Some of the best targets for antibodies are cell surface receptors involved in signal transduction pathways. Several of these pathways are deregulated in cancer and other diseases, and have been the target of chemical drugs for a number of years. However, the efficacy of some chemical inhibitors has been hindered by mutation-acquired resistance of the target proteins. Targeting these receptors with antibodies may have several advantages. Antibodies not only work by directly inhibiting (or in some cases, stimulating) the target protein, they also trigger antibody-dependent cell toxicity and in come cases, apoptosis of the target cell. Moreover, synthetic antibody technology allows the generation of antibodies to multiple epitopes or multiple conformations of a protein or protein complex, such that affinity and specificity, as well as efficacy, are increased (Gao et al, 2009). The ability to generate several different antibodies with different affinity profiles against a single antigen also allows fine-tuned modulations of a signaling pathway, broadening the usefulness of these antibodies from therapeutic candidates to outstanding research reagents that will increase our understanding of the pathways.
We have targeted a number of signaling pathway components with synthetic antibodies, including HER2 (Fisher et al, 2010; Miller et al, 2012; Owen et al, 2013), DR5 (Li et al, 2006), VEGF (Sidhu et al, 2004; Fellouse et al, 2004; Fellouse et al, 2005(1)(2); Fellouse et al, 2007), the erythropoeitin receptor (Hu et al, 2013), and the KIT receptor kinase (Reshetnyak et al, 2013). We are also involved in a number of projects geared toward the use of synthetic antibodies to fight infectious diseases caused by Staphyloccus aureus bacteria (Karauzum et al, 2012) and Ebola virus (Chen et al, 2014; Koellhoffer et al, 2012). Additionally, we have been working with Drs. Raymond Reilly and Mitchell Winnik to generate antibodies with enhanced therapeutic effectiveness by chemical linkage to cytotoxic radionuclides (Liu et al, 2014).
Besides modulation of signaling pathways and therapeutic potential, antibodies have many other uses in biological research. For example, we have collaborated with several groups to use antibodies as chaperones during the crystallization of proteins and RNA for structural analysis (Ye et al, 2007; Uysal et al, 2009; Koldobskaya et al, 2011). Our antibodies have also been used as tools to study protein and RNA structure and function (Ye et al, 2007; Newton et al, 2008; Laver et al, 2012; Zalatan et al, 2012), and to stabilize protein complexes during functional experiments (Shukla et al, 2014).
