Engineering and structural characterization of cross-reactive, pH-sensitive antibodies against the long-chain α-neurotoxins

Jack Wade*

*Corresponding author for this work

Research output: Book/ReportPh.D. thesis

246 Downloads (Pure)

Abstract

Antibodies are proteins secreted by immune cells during the immune response to a pathogen. By binding to the surfaces of pathogens with high affinity and specificity, antibodies both neutralize pathogens and coordinate the immune response. To generate an effective antibody immune response, the immune system samples over 108 unique antibodies in the naïve antibody repertoire, which expands exponentially following a pathogenic challenge. For over a century, the treatment of snakebite envenomings has taken advantage of the ability of the immune system to generate highly potent, broadly neutralizing antibodies by immunizing large animals with snake venoms and extracting the antibodies for treatment. This forms modern-day antivenom and validates the capability of antibodies to neutralize the venom of different snakes. However, as the antibodies are of non-human origin, they have an increased likelihood of causing adverse reactions when used to treat snakebite victims due to their heterologous nature, which is associated with high immunogenicity. Therefore, developing an antivenom that consists of human antibodies is one approach to improving snakebite treatment.
Antibodies are both highly potent and specific for their target and are heavily used as therapeutics, accounting for over half of the approved biopharmaceuticals in the last 4 years. Beyond their high affinity and specificity, antibodies have been engineered to have different mechanisms of action. For example, a generation of antibodies has been developed to bind non-stoichiometrically, to neutralize more than one target molecule in their lifetime, allowing them to be administered at a lower dose than a conventional antibody. These antibodies require their binding affinity to be pH-dependent, specifically to have a low affinity at acidic pH, allowing the antibodies to release antigens for lysosomal degradation whilst being exposed to low pH during cellular recycling. As a result, antibodies are returned to the bloodstream unbound and ready to bind to another target molecule at neutral pH. This property could potentially help lower the dose and economic feasibility of a more efficacious antivenom for developing countries affected by snakebite. Further, developing antibodies that are cross-reactive, and therefore can bind to structurally similar toxins in different snakes, would broaden the usage (polyvalence) of a prospective antivenom and lower the number of antibodies required for treatment.
This thesis focuses on the development of neutralizing human monoclonal antibodies against long-chain α-neurotoxins from elapid snakes and understanding the molecular determinants underpinning pH-dependent antigen binding in antibodies. We began by discovering antibodies from a naïve human antibody phage display library and optimizing them for improved cross-reactivity through light chain shuffling. By doing so, we enhanced the neutralization capacity of one antibody in vivo, showcasing the use of phage display technology and light chain shuffling as an approach to discovering broadly neutralizing antibodies against this toxin family (Chapter 4).
Building further upon the abovementioned study, we report the crystal structure of one of these antibodies bound to a long-chain α-neurotoxin, determined at 1.6 Å, and elucidate the basis for the neutralization mechanism of this lineage of antibodies. Through the antibody heavy chain complementary determining region 3, these antibodies mimicked conserved interactions between long-chain α-neurotoxins and the acetylcholine receptor to neutralize long-chain α-neurotoxins and achieve broad cross-reactivity. This antibody also bound pH-dependently to all long-chain α-neurotoxins tested, initiating further structural studies to investigate the pH-dependent binding mechanism. Determining the structures of the antibody bound to long-chain α-neurotoxin at different pH identified a network of residues that respond in concert to low pH in the antibody structure, located at the interface between the antibody heavy and light chain (Chapter 5). Thus, these data provide an indication of a paratope-independent, pH-sensitive binding mechanism.
The discovery of pH-dependent antibodies is low throughput and necessitates multiple engineering steps and discovery campaigns to tune the antibody affinity at both neutral and acidic pH. This thesis aimed to improve the discovery of pH-dependent antibodies using in vitro display technology. We had observed that pH-dependent antigen binding could be encoded exclusively away from the paratope, potentially in the heavy-light chain interface, and used this insight to conceptualize an approach to engineer a generic pH switch into the antibody variable domain. We designed and validated a phage display library targeting the antibody heavy-light chain interface framework region to introduce a generic pH switch into the antibody variable domain. With pH-dependent binding engineered as a pre-determined feature into antibodies, we envisage a high throughput approach to discovering and/or engineering pH-dependent antibodies using in vitro display technologies a priori (Chapter 6).
Lastly, we employed a self-assembly protein domain to expand the valence and neutralization capacity of nanobodies targeting long-chain α-neurotoxins and validated parameters important for their application (Chapter 7). We produced stable, multivalent nanobody-based proteins engineered to contain up to sixteen binding domains that display enhanced neutralization potency of long-chain α-neurotoxins in vitro, and were able to be recycled in a cellular assay when engineered to contain IgG-Fc.
Overall, this work defines the molecular determinants of antibody recognition and neutralization of long-chain α-neurotoxins, which will be useful to guide the engineering of these antibodies to improve their neutralization potency for use in a recombinant antivenom. I have also developed an approach to enhance the neutralization capacity of nanobodies and expand their half-life and effector function properties, which might facilitate the use of nanobodies in the treatment of different diseases. Furthermore, the molecular insights gained through the work behind this thesis on pH-dependent antigen binding can be used to guide the engineering of this property into antibodies. This will have utility in improving the pharmacology of antibodies that suffer from target-mediated clearance, improve their cost-effectiveness by having a longer duration of action and a lower administration dose, and potentially yield therapies with better patient compliance (due to more infrequent dosing). Lastly, the insights acquired into how antibodies bind pH-dependently to their target due to properties lying outside of the paratope and epitope interface may have many further implications in the de novo discovery of antibodies with pH-dependent binding properties. One example includes the design of specialized in vitro display antibody libraries with pre-established pH-dependent binding properties that could be used in a generic fashion to discover recycling antibodies and/or antibodies that engage with their target antigen at exact anatomical locations at the right time. My hope is that this will one day lead to more efficient development of therapies against snakebite envenoming, but also beyond in areas such as oncology, infectious diseases, and autoimmune diseases.
Original languageEnglish
Place of PublicationKgs. Lyngby, Denmark
PublisherDTU Bioengineering
Number of pages146
Publication statusPublished - 2023

Fingerprint

Dive into the research topics of 'Engineering and structural characterization of cross-reactive, pH-sensitive antibodies against the long-chain α-neurotoxins'. Together they form a unique fingerprint.

Cite this