Discovery of antibodies with pH-dependent binding towards snake venom toxin

Snakebite envenoming is a neglected tropical disease that kills 81,000 to 138,000 people each year and leaves many more with permanent sequelae. The only specific treatment for snakebite envenoming relies on animal-derived (traditional) antivenoms. Although these antivenoms have saved countless lives, they are plagued with several drawbacks, such as eliciting adverse reactions, a requirement for high doses, and low therapeutic content. Consequently, researchers are working to develop alternative treatment strategies. Prominently featuring among them are oligoclonal recombinant antivenoms, which are based on mixtures of recombinantly produced (human) monoclonal antibodies targeting medically relevant snake venom toxins. Since such antivenoms will contain antibodies that are more compatible with the human immune system, they will have a lower risk of eliciting adverse reactions in the patients. Plus, with high therapeutic content, they are likely to require lower treatment doses than traditional antivenoms. Additionally, the cost of manufacturing oligoclonal recombinant antivenoms is estimated to be competitive or even lower in the future compared to traditional antivenoms, which may further reduce treatment costs. Indeed, both reduced dosing and cost are crucial for the successful implementation of recombinant antivenoms. A potential solution to achieving lower dosing requirements, and consequently reduced treatment costs, can be achieved by the application of monoclonal antibodies that bind their cognate antigen in a pH-dependent manner, also known as recycling antibodies. Such antibodies bind antigens at the neutral pH in the bloodstream and dissociate from them in the acidic environment of endosomes. This allows the antibodies to be recycled back into circulation via neonatal Fc receptor (FcRn)-binding, while the antigen is destined for lysosomal degradation. Thus, compared to a non-recycling antibody, that binds antigen only once per binding site, a single recycling antibody can bind and eliminate multiple antigens in its lifetime. This enables recycling antibodies to achieve therapeutic effects at lower doses than non-recycling antibodies, and may thus find utility for the development of low-cost recombinant antivenoms.
In this thesis, I have worked on discovering antibodies that target snake toxins in a pH-dependent manner using phage display technology. While histidine doping in the variable domains of antibodies is the most used strategy to introduce pH-dependent binding, I explored whether pH-dependent antigen-binding antibodies can be discovered directly from a naïve human antibody library that consists of naturally occurring variable domains. Indeed, an antibody with pH dependent binding towards the α-cobratoxin (α-cbtx) was discovered and found to be entirely devoid of histidines in its paratope. The findings of the study demonstrate that pH-dependent antigen-binding antibodies can be discovered from antibody libraries with naturally occurring variable domains, and that pH-dependent interactions can, at least at times, be derived from non-histidine amino acid residues.
Next, it was investigated whether light-chain shuffling can be employed as a strategy to enhance the pH-dependent antigen-binding properties of pre-existing antibodies. Light-chain shuffled libraries of two myotoxin-II (M-II)-targeting antibodies and one α-cbtx-targeting antibody were generated and coupled to phage display for selection of pH-dependent antigen-binding antibodies. This resulted in the discovery of an M-II- and an α-cbtx-targeting pH-dependent antibody, thus demonstrating the utility of the employed strategy.
For efficient performance of antibodies with pH-dependent antigen-binding properties, it is crucial that they are efficiently rescued via FcRn binding. However, it has been found that biophysical properties of the fragment antigen-binding region (Fab) and antigen-binding can influence the FcRn-mediated rescue of antibodies. Thus, the discovered pH-dependent antigen-binding antibodies were further studied for their interaction with FcRn and their behavior in a cellular assay, which revealed distinct effects of Fab and antigen-binding on the cellular transport properties of the antibodies. Finally, by employing cellular assays, the study demonstrated that the discovered antibodies with pH-dependent antigen-binding properties performed as recycling antibodies.
While recycling antibodies against snake venom toxins have been discovered in this project, it is important to note that snakebite envenoming involves complex toxicokinetics, and the utility of recycling antibodies in such a setting is not known. Nevertheless, this PhD thesis explored antibody libraries with natural domains, light-chain shuffling, phage display selection strategies, and cellular assays for the discovery and validation of pH- dependent antigen-binding antibodies. Additionally, the role of Fab and antigen-binding on the cellular transport properties of these antibodies were explored, which provided insights into factors that could affect their performance as recycling antibodies. Thus, it is my hope that the studies presented in this thesis can enable further understanding of how to discover and engineer pH-dependent antigen-binding antibodies, and that the methodologies developed in the work behind this thesis can find applications in snakebite envenoming as well as other indications, such as oncology, autoimmunity, infectious diseases, and general protein science.