(191cm) Selection and Affinity Enhancement of Alpha-Synuclein-Specific Single Domain Antibody Using Experimental and Simulation Techniques

The ability to modulate protein-protein interactions is of great interest to both fundamental biological science and applied research such as drug design. While many studies have demonstrated the ability to use structure-based information to rationally improve protein-protein interactions, the rational design of target-specific binders is still an unsolved problem. In this context, single domain antibodies (VHHs) are promising designable molecules, given their small size, stability, and ability to bind surfaces of any shape. Alpha-Synuclein (AS) is a natively disordered protein, implicated in the pathogenesis of Parkinson’s disease (PD) and related neurological disorders. The Non-Amyloid Component (NAC) region of AS has been previously targeted for inhibiting the aggregation and/or reducing toxicity of AS. Various methods such as yeast surface display, phage display and, more recently, rational design approaches have been used for generating antibody fragments for this purpose. In this work, we have pursued a dual modeling-experimental approach for rationally designing an AS-specific VHH.

Starting from an immunized Camelid library against the NAC region of A53T mutant of AS (A53T), a single-round of a bacteria-based selection technique was used to obtain a NAC-specific VHH. Atomistic models of the VHH and the VHH-NAC binding were constructed using a combination of Molecular Dynamics techniques (Replica Exchange Molecular Dynamics, Umbrella Sampling) and numerical calculations (Finite-Difference Poisson-Boltzmann Equation). Furthermore, affinity enhancing mutations based on computational models were proposed and tested using experimental techniques such as ELISA and Surface Plasmon Resonance (SPR). Using this dual approach of prediction and experimental verification, we were able to enhance the affinity of the VHH by an order of magnitude. Hence, our work demonstrates the ability to develop a predictive in-silico model of binding for the VHH-NAC system. This approach can be especially powerful for targeting natively disordered and weakly immunogenic antigens such as AS, for which, crystal structures are not readily available and experimentally validated computational models can be leveraged for rational design and affinity enhancement.