Decoding the Structure–Function Correlation of Adeno-Associated Virus 2 Capsid Mutants Recognition by A20 Antibody: A Predictive Modeling Using Coarse-Grained Simulations

Adeno-associated virus serotype 2 (AAV2) is widely used as a gene therapy vector due to its favorable safety profile and broader transduction capabilities. However, pre-existing immunity poses a significant barrier to its therapeutic applications. In this study, we employed coarse-grained elastic network molecular dynamics simulations to investigate the structural and conformational dynamics of the wild-type AAV2 capsid and its six capsid variants (Q263A, S264A, S384A, Q385A, V708A, and V708K) upon binding to a mouse monoclonal antibody (A20), a robustly used AAV2-specific antibody. Notably, A20 recognizes a few immunodominant epitopes that can be utilized to design AAV2 mutants with robust resistance to human neutralizing sera. Our analysis revealed that the involvement of three different symmetry-related subunits of the AAV2 capsid is critical in mediating interactions with A20, particularly through its heavy-chain complementarity-determining regions (CDRs). Per-residue energy decomposition analysis identified key interaction hotspots, which are in agreement with the experimental neutralization data for escape mutants. Structural descriptors, such as root-mean-square deviation (RMSD), radius of gyration (R_g), solvent-accessible surface area (SASA), center-of-mass (COM) distances, and contact probabilities, were well correlated with experimental A20 binding data. A predictive model was developed using multiple linear regression (R_{CrossValidation}² = 0.949), successfully capturing the relationship between mutation-induced structural changes in AAV2 and fold reduction in A20 binding affinities. This integrative approach provides mechanistic insights into capsid-antibody recognition and offers a structure-guided, rational framework for designing AAV2 variants with reduced immunogenicity, thereby advancing the development of next-generation gene therapy vectors.