Antibodies specific for different conformers (monomers, oligomers or fibrils) of aggregation-prone polypeptides associated with neurodegenerative diseases (e.g. Alzheimer’s and Parkinson’s disease) have a wide range of therapeutic and diagnostic applications. Nevertheless, generating such antibodies is challenging because of the complexity of the antigens, the limited control over antigen presentation afforded by immunization, and the difficulty in obtaining antibody variants with both conformational and sequence specificity. To overcome such limitations, we are developing methods for designing lead antibody fragments by mimicking natural homotypic peptide interactions that mediate amyloid formation. Our approach involves grafting key recognition peptides into the binding loops of antibody fragments such as human variable (V<sub>H</sub>) domains and single-chain variable fragments (scFvs). However, these lead antibody fragments have modest (high nanomolar) affinities. To improve their binding affinity, we are using random and focused mutagenesis as well as yeast surface display to select variants with improved properties. Interestingly, traditional yeast surface display yields antibody fragments with improved binding affinity and reduced stability. Nevertheless, we are developing multiple approaches to improve antibody binding affinity while maintaining stability. These approaches include co-selection for antigen binding and antibody stability using conformational ligands as well as selection of antibody fragments in highly stringent solution conditions that suppress non-specific binding. Our findings highlight the importance of co-optimizing multiple antibody properties to improve rapid selection of high-quality antibody fragments.

Improved methods for co-evolving the affinity and stability of antibody fragments specific for amyloid-forming polypeptides

251st National Meeting (2016)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.