Periodic Reporting for period 2 - MABSTER (Monoclonal Antibodies with Binding Sensitive To Environmental Regulation)
Okres sprawozdawczy: 2021-07-01 do 2022-12-31
Snakebite envenoming is a Neglected Tropical Disease (NTD) that each year affects 2.5 million victims and kills >100,000, unless they are treated with antivenom. Conventional antivenoms, derived from immunized animals, inflict serum sickness and anaphylaxis in patients, and are costly to manufacture.
Monoclonal human antibodies with special toxin-binding properties that are sensitive towards regulation by their microenvironment, which may be discovered using phage display selection, may solve this issue, providing significant societal impact by enabling the development of cost-effective antivenoms to victims in low and middle-income countries.
In this project, phage display selection, high-density peptide microarray technology, and antibody engineering techniques will in three scientific objectives be harnessed in the pursuit of developing novel methodologies for discovery of therapeutic human monoclonal antibodies that are recyclable (can neutralize more than one snake toxin per antibody), broadly cross-reactive (can neutralize different types of snake toxins), and that are both broadly cross-reactive and recyclable at the same time. This will open up for entirely new ways of designing biotherapeutics against complex indications, such as snakebite envenoming, but also cancer, infectious, and parasitic diseases, where the targets can be elusive due to hyper-mutability.
Monoclonal human antibodies with special toxin-binding properties that are sensitive towards regulation by their microenvironment, which may be discovered using phage display selection, may solve this issue, providing significant societal impact by enabling the development of cost-effective antivenoms to victims in low and middle-income countries.
In this project, phage display selection, high-density peptide microarray technology, and antibody engineering techniques will in three scientific objectives be harnessed in the pursuit of developing novel methodologies for discovery of therapeutic human monoclonal antibodies that are recyclable (can neutralize more than one snake toxin per antibody), broadly cross-reactive (can neutralize different types of snake toxins), and that are both broadly cross-reactive and recyclable at the same time. This will open up for entirely new ways of designing biotherapeutics against complex indications, such as snakebite envenoming, but also cancer, infectious, and parasitic diseases, where the targets can be elusive due to hyper-mutability.
In this project, we have made progress in several different thematic areas, presented below:
Synthetic biology toolbox
To enable multiple different projects focusing on developing broadly-neutralizing and/or pH-sensitive antibodies, we have developed a large range of synthetic biology tools and protocols. These include:
1. Construction of 8 expression vectors for E. coli with different combinations of solubilization and biotinylation tags.
2. Construction of 3 expression vector for P. pastoris for secreted expression.
3. Cloning of 22 snake toxins into those expression vectors:
2.a 7 native toxins from snakes from South America and Sub-saharan Africa.
2.b 15 consensus toxins (long-chain and short-chain 3FTxs, and Kunitz-type serine protease inhibitors) from Sub-saharan Africa and Asia.
3. Expression of some of the toxins in E. coli and validation of their functionality with a patch-clamp assay on cells expressing muscle-type nicotinic AChR.
Genomics, transcriptomics, expression of dormant genes
Studying venom and the evolutionary trends is critical to develop antivenoms and neutralise the most relevant toxins. Therefore, as part of a collaboration with the Beijing genomics institute, we have assembled the genome of the European Nose-horned Viper (Vipera ammodytes ammodytes) and annotated the toxins backed on transciptomics and proteomics analysis.
The combination of multiple big-data techniques allows us to discover the sequence of toxins at the genome level, that can not be detected at the mRNA or protein level, and therefore are imposible to obtain from the natural source. Among the dormant toxins identified, we have chosen a toxin with potential biotechnological application, (an L-amino acid oxidase - LAAO) as a proof of concept for the potential unlocked by big data techniques. We have cloned, produced and purified the dormant LAAO. We are pending activity assays to confirm its functionality.
Designed toxins and their use as antigens for discovering broadly-neutralizing antibodies
In one area, we focused on developing techniques for finding broadly-neutralizing antibodies through phage display. For that purpose, we developed a tool for the design of special immunogens that can be used to discover broadly-neutralizing antibodies. We designed immunogens based on scorpion and spider toxins using the tool, and expressed them in bacteria and yeast cells. We pursued a phage display campaign against the designed immunogens, discovering antibodies that can cross-recognise natural toxins from entirely different orders (Scorpiones and Aranae). We published a review (Rivera-de-Torre et.al 2022) and drafted a manuscript about the successful use of designed immunogens.
We have also performed phage display selection campaigns on 3 native snake toxins and 5 consensus snake toxins, expressed and evaluated the best toxin binders using BLI, and identified a common binding motif.
Finally, we received funding from ERC-PoC to develop the first black widow antivenom based on a monoclonal antibody and started patenting processes for both this antibody, as well as an entirely different subproject focused on allergy (the latter pending the decision on a new ERC-PoC grant application).
To further understand the limits of consensus toxins as tool to discovery broadly-neutralising antibodies, we have expanded the project to black widow spider projects. We have designed latrodectins (small molecular weight peptides present onf black widow spider venoms) and we aim to use them as target antigens in a phage display campaign.
Discovery and design of pH-sensing antibodies
We have developed surfaced-based biosensor assays (BLI and SPR) to identify cross-reactive antibodies that display pH-sensitive kinetics against long-neurotoxins. To understand the molecular basis of pH-sensitive binding, a crystallography pipeline has been set up, involving improving microbial production of the variable domains of antibodies and implementing gel filtration purification methods to obtain pure bound scFv antibodies. Resultant data analysis of a structure of an antibody in complex with a long-chain neurotoxin has revealed the mechanism underpinning cross-reactivity and neutralization. We have set up optimization screens at variable pHs to understand pH-sensitivity. Data has been acquired for over 50 crystals obtained at 5 different pH levels. To date, structures are being determined at comparable resolutions for each pH. Combined, this allows us to study and unravel molecular mechanisms for pH-sensitive antibody-antigen interactions.
In a parallel subproject, we have developed phage display-based protocols that have allowed us to identify a pH sensitive scFv against alpha-cobratoxin from a naïve scFv library. Moreover, an anti-myotoxin II scFv with low pH-sensitivity was used to create a light-chain shuffled library, from which a more pH sensitive scFvs were successfully discovered using novel phage display pH selections strategies. The scFvs were reformatted to IgGs and Fabs, and their binding kinetics have been characterized. This has provided us with robust methodologies and data that now allow us to routinely discover pH sensitive antibodies and study their binding behaviours. The discovered pH-sensitive antibodies have further been characterized to their pH-sensitive antigen binding properties in a cellular assay (Human endothelial cell-based recycling assay: HERA). This data from HERA has provided the indications the recycling nature of these antibody.
Structural biology
To perform in-depth studies of antibodies with special binding properties (broadly-neutralizing and/or pH-sensitive binding), we have established crystallographic screening protocols and selected buffer screening conditions for co-crystallization experiments. We have used these protocols to study apparent nucleation events and crystal growth for toxin-antibody complexes. We have further performed crystallographic model building and refinement and determined composition of asymmetric unit to confirm successful co-crystallization of one antigen-antibody complex at satisfactory resolution. Finally, we have performed crystal structure analysis, including delineating composition of intermolecular interfaces, confirming stoichiometry and generated hypotheses for mechanisms underlying mode of action and response to environmental factors.
Combined, the above describes areas of progress have enabled us to create the tools, protocols, and pipelines that we can now use to further achieve the Sceintific Objectives in this ERC StG project. We have thus sown the seeds for several projects, and are well on our way to reap the fruits in the final stages of the project. A decent number of scientific manuscripts have already been published, but we expect the bulk of the work performed will be published within 2023.
Synthetic biology toolbox
To enable multiple different projects focusing on developing broadly-neutralizing and/or pH-sensitive antibodies, we have developed a large range of synthetic biology tools and protocols. These include:
1. Construction of 8 expression vectors for E. coli with different combinations of solubilization and biotinylation tags.
2. Construction of 3 expression vector for P. pastoris for secreted expression.
3. Cloning of 22 snake toxins into those expression vectors:
2.a 7 native toxins from snakes from South America and Sub-saharan Africa.
2.b 15 consensus toxins (long-chain and short-chain 3FTxs, and Kunitz-type serine protease inhibitors) from Sub-saharan Africa and Asia.
3. Expression of some of the toxins in E. coli and validation of their functionality with a patch-clamp assay on cells expressing muscle-type nicotinic AChR.
Genomics, transcriptomics, expression of dormant genes
Studying venom and the evolutionary trends is critical to develop antivenoms and neutralise the most relevant toxins. Therefore, as part of a collaboration with the Beijing genomics institute, we have assembled the genome of the European Nose-horned Viper (Vipera ammodytes ammodytes) and annotated the toxins backed on transciptomics and proteomics analysis.
The combination of multiple big-data techniques allows us to discover the sequence of toxins at the genome level, that can not be detected at the mRNA or protein level, and therefore are imposible to obtain from the natural source. Among the dormant toxins identified, we have chosen a toxin with potential biotechnological application, (an L-amino acid oxidase - LAAO) as a proof of concept for the potential unlocked by big data techniques. We have cloned, produced and purified the dormant LAAO. We are pending activity assays to confirm its functionality.
Designed toxins and their use as antigens for discovering broadly-neutralizing antibodies
In one area, we focused on developing techniques for finding broadly-neutralizing antibodies through phage display. For that purpose, we developed a tool for the design of special immunogens that can be used to discover broadly-neutralizing antibodies. We designed immunogens based on scorpion and spider toxins using the tool, and expressed them in bacteria and yeast cells. We pursued a phage display campaign against the designed immunogens, discovering antibodies that can cross-recognise natural toxins from entirely different orders (Scorpiones and Aranae). We published a review (Rivera-de-Torre et.al 2022) and drafted a manuscript about the successful use of designed immunogens.
We have also performed phage display selection campaigns on 3 native snake toxins and 5 consensus snake toxins, expressed and evaluated the best toxin binders using BLI, and identified a common binding motif.
Finally, we received funding from ERC-PoC to develop the first black widow antivenom based on a monoclonal antibody and started patenting processes for both this antibody, as well as an entirely different subproject focused on allergy (the latter pending the decision on a new ERC-PoC grant application).
To further understand the limits of consensus toxins as tool to discovery broadly-neutralising antibodies, we have expanded the project to black widow spider projects. We have designed latrodectins (small molecular weight peptides present onf black widow spider venoms) and we aim to use them as target antigens in a phage display campaign.
Discovery and design of pH-sensing antibodies
We have developed surfaced-based biosensor assays (BLI and SPR) to identify cross-reactive antibodies that display pH-sensitive kinetics against long-neurotoxins. To understand the molecular basis of pH-sensitive binding, a crystallography pipeline has been set up, involving improving microbial production of the variable domains of antibodies and implementing gel filtration purification methods to obtain pure bound scFv antibodies. Resultant data analysis of a structure of an antibody in complex with a long-chain neurotoxin has revealed the mechanism underpinning cross-reactivity and neutralization. We have set up optimization screens at variable pHs to understand pH-sensitivity. Data has been acquired for over 50 crystals obtained at 5 different pH levels. To date, structures are being determined at comparable resolutions for each pH. Combined, this allows us to study and unravel molecular mechanisms for pH-sensitive antibody-antigen interactions.
In a parallel subproject, we have developed phage display-based protocols that have allowed us to identify a pH sensitive scFv against alpha-cobratoxin from a naïve scFv library. Moreover, an anti-myotoxin II scFv with low pH-sensitivity was used to create a light-chain shuffled library, from which a more pH sensitive scFvs were successfully discovered using novel phage display pH selections strategies. The scFvs were reformatted to IgGs and Fabs, and their binding kinetics have been characterized. This has provided us with robust methodologies and data that now allow us to routinely discover pH sensitive antibodies and study their binding behaviours. The discovered pH-sensitive antibodies have further been characterized to their pH-sensitive antigen binding properties in a cellular assay (Human endothelial cell-based recycling assay: HERA). This data from HERA has provided the indications the recycling nature of these antibody.
Structural biology
To perform in-depth studies of antibodies with special binding properties (broadly-neutralizing and/or pH-sensitive binding), we have established crystallographic screening protocols and selected buffer screening conditions for co-crystallization experiments. We have used these protocols to study apparent nucleation events and crystal growth for toxin-antibody complexes. We have further performed crystallographic model building and refinement and determined composition of asymmetric unit to confirm successful co-crystallization of one antigen-antibody complex at satisfactory resolution. Finally, we have performed crystal structure analysis, including delineating composition of intermolecular interfaces, confirming stoichiometry and generated hypotheses for mechanisms underlying mode of action and response to environmental factors.
Combined, the above describes areas of progress have enabled us to create the tools, protocols, and pipelines that we can now use to further achieve the Sceintific Objectives in this ERC StG project. We have thus sown the seeds for several projects, and are well on our way to reap the fruits in the final stages of the project. A decent number of scientific manuscripts have already been published, but we expect the bulk of the work performed will be published within 2023.
In this project, we have made progress beyond state of the art in several different thematic areas, presented below together with the expected results going forward:
Designer toxins and their use as antigens for discovering broadly-neutralizing antibodies
We have, for the first time, used designed immunogens (e.g. consensus toxins) in a phage display-based antibody discovery campaign and demonstrated that we can discover antibodies targeting structural epitopes (something that was expected by us, but which was deemed impossible by leading researchers in the field). We plan to apply the design tool to different toxin families to showcase the potency of the tool to obtain broadly-neutralising antibodies. We expect to publish at least two new publications from the activities developed in these research lines and obtain further funding for commercialization and transfer of the research results. Moreover, we have, for the first time, used the CyDisCo expression system with snake toxins.
Discovery and design of pH-sensing antibodies
We have obtained more and better cross-reactive binders by using differently expressed toxins for the phage display selections: yeast and mammalian expression systems are on the way. We have also used our established protocols for finding more good pH-sensitive and cross-reactive antibodies and study their binding properties to unravel the fundamental mechanisms involved in pH-sensitive antigen binding and broadly-neutralizing properties. In this regard, we have characterized Fc and FcRn binding properties of selected IgGs at pH 7.4 and pH 5.4 using ELISA, and we have employed a human endothelial recycling assay (HERA) to investigate the recycling properties of the pH-sensitive and non-pH-sensitive IgGs with and without the toxin. Preliminary results indicate that we have established a robust pipeline for designing antibodies with special binding properties.
Structural biology
Currently, structural understanding of antibody-mediated toxin neutralization is close to completely absent in the general field of (animal) toxinology. We have obtained deep understanding of toxin cross-neutralization using anti-long-neurotoxin antibodies by establishing a crystallography pipeline that can generate atomic-level insights into epitope-paratope interaction at high throughput. Using this pipeline, our crystallography results reveal a conformational epitope consisting of conserved residues important in the toxin interacting with the cognate acetyl choline receptor (AChR). We have observed that some of our potent antibodies mimic residues on the nAChR targeted by the conserved toxin residues, collectively showing the basis for neutralization being born through an element of structural mimicry. Despite many of these core interactions being shared in related antibodies, there are marked differences in pH-sensitive binding to different neurotoxins. This has led us to develop a surprising hypothesis around toxin neutralization, which may be mediated by an allosteric mechanism communicated by the antibody light chain to introduce pH-sensitive binding. In early insights, we observed that the antibody structure is responsive to changes in pH at the interface between the light and heavy chains. This mechanism may possibly be ”universal”, which could mean that we have cracked one of the codes for making universal pH-sensing antibody scaffolds. By end of the project, we will also have an integrative framework explaining biochemical and biophysical characteristics of broadly-neutralizing candidate antibodies.
Designer toxins and their use as antigens for discovering broadly-neutralizing antibodies
We have, for the first time, used designed immunogens (e.g. consensus toxins) in a phage display-based antibody discovery campaign and demonstrated that we can discover antibodies targeting structural epitopes (something that was expected by us, but which was deemed impossible by leading researchers in the field). We plan to apply the design tool to different toxin families to showcase the potency of the tool to obtain broadly-neutralising antibodies. We expect to publish at least two new publications from the activities developed in these research lines and obtain further funding for commercialization and transfer of the research results. Moreover, we have, for the first time, used the CyDisCo expression system with snake toxins.
Discovery and design of pH-sensing antibodies
We have obtained more and better cross-reactive binders by using differently expressed toxins for the phage display selections: yeast and mammalian expression systems are on the way. We have also used our established protocols for finding more good pH-sensitive and cross-reactive antibodies and study their binding properties to unravel the fundamental mechanisms involved in pH-sensitive antigen binding and broadly-neutralizing properties. In this regard, we have characterized Fc and FcRn binding properties of selected IgGs at pH 7.4 and pH 5.4 using ELISA, and we have employed a human endothelial recycling assay (HERA) to investigate the recycling properties of the pH-sensitive and non-pH-sensitive IgGs with and without the toxin. Preliminary results indicate that we have established a robust pipeline for designing antibodies with special binding properties.
Structural biology
Currently, structural understanding of antibody-mediated toxin neutralization is close to completely absent in the general field of (animal) toxinology. We have obtained deep understanding of toxin cross-neutralization using anti-long-neurotoxin antibodies by establishing a crystallography pipeline that can generate atomic-level insights into epitope-paratope interaction at high throughput. Using this pipeline, our crystallography results reveal a conformational epitope consisting of conserved residues important in the toxin interacting with the cognate acetyl choline receptor (AChR). We have observed that some of our potent antibodies mimic residues on the nAChR targeted by the conserved toxin residues, collectively showing the basis for neutralization being born through an element of structural mimicry. Despite many of these core interactions being shared in related antibodies, there are marked differences in pH-sensitive binding to different neurotoxins. This has led us to develop a surprising hypothesis around toxin neutralization, which may be mediated by an allosteric mechanism communicated by the antibody light chain to introduce pH-sensitive binding. In early insights, we observed that the antibody structure is responsive to changes in pH at the interface between the light and heavy chains. This mechanism may possibly be ”universal”, which could mean that we have cracked one of the codes for making universal pH-sensing antibody scaffolds. By end of the project, we will also have an integrative framework explaining biochemical and biophysical characteristics of broadly-neutralizing candidate antibodies.