Engineered vectors for single-domain antibody selection from immune and synthetic libraries

The revolutionary and serendipitous discovery of heavy chain antibodies in the blood of camelids in 1993 aroused great interest. This discovery has marked a breakthrough in the field of antibodies because the variable domain of this class of antibodies, known as VHH or nanobody or single-domain antibodies, has unique properties. This variable domain is very small (~15 kDa) compared to conventional antibodies (~150 kDa), it is soluble due to the absence of exposed hydrophobic amino acid residues, and it can be expressed individually in bacteria as a recombinant protein while retaining target affinity comparable to that of the whole antibody. These distinctive properties have made single-domain antibodies a versatile tool with a wide range of applications, from basic research to diagnostics and therapy. During my PhD, I focused on engineering vectors suitable for displaying nanobodies on phages and bacteria. I systematically tested the functionality of these vectors by generating libraries encompassing both synthetic and immune nanobodies. To get this research project started, I chose an efficient cloning system that could guarantee the creation of highly complex libraries: Gateway® technology. Commercially available vectors for ligand selection by phage display do not support the use of this cloning method and, as they are mostly phagemid-based, require long times to perform repeated selection steps. I strategically modified an existing phage vector, already adapted for phage display technology, to take advantages of Gateway cloning. This cloning system enabled the construction of a phage library of synthetic humanized nanobodies, using a synthetic sequence encoding a nanobody with a humanized scaffold framework (hCFW) as a template. Despite achieving a sufficiently complex library and enriching it with nanobodies targeting the thioredoxin protein from the bacterium Pyrococcus furiosus (PfTrx), subsequent attempts to isolate individual clones specifically directed against the target proved challenging. This limitation may be due to the low binding affinity of the nanobodies in the initial library, which probably requires the inclusion of maturation steps. The destination genetic cassette (DEST) proposed by Invitrogen for modifying vectors used with the Gateway cloning system is unsuitable for classical phagemid vectors used in phage display. Counter-selection of the destination cassette, which contains the gene encoding the toxin CcdB, proves ineffective in the bacterial strains used for phage display beacause they also carry the gene for the antitoxin (CcdA) on the F plasmid, which can neutralise the toxin. To overcome this issue, I designed and implemented the modification of a phagemid vector by mutagenizing the classic destination cassette. This modification allowed overexpression of the ccdB gene upon induction with the IPTG. Under toxin overexpression conditions, the resulting vector facilitated the selection of recombinant clones, even in normally resistant bacterial strains. To validate the functionality of this new vector, I constructed an immune nanobody library from a llama vaccinated with PfTrx and isolated a specific ligand for this target protein. Finally, I investigated the potential of the bacterial display system by evaluating three different Escherichia coli surface proteins for their suitability in this selection process. I prepared three expression vectors capable of producing and transporting the three proteins to the bacterial outer membrane. These proteins were fused to a passenger protein, specifically a domain of the FbaB protein from Streptococcus pyogenes (SpyCatcher). This setup allowed for the evaluation of the effectiveness of the three autotransporters in delivering a foreign protein to the outer membrane of E. coli using the SpyTag-SpyCatcher binding system. After identifying the most effective autotransporter, I transformed the corresponding expression vector into a destination vector by incorporating our modified cassette that overexpressing the CcdB toxin to ensure compatibility with the Gateway technology. Using this Gateway cloning approach, I demonstrated the ability of the vector to display a specific nanobody on the bacterial surface. The next step is to use this vector to generate a nanobody library and perform selection through bacterial display methodology.