Integrated computational and experimental immunoengineering of Adeno-associated virus capsid T cell epitopes in mice

Adeno-associated virus (AAV) vectors are widely used in gene therapy, but their immunogenicity remains a significant challenge, limiting long-term efficacy and the feasibility of repeated administration. In this study, we used a combined approach of computational prediction and experimental validation to engineer AAV9 capsids with reduced immunogenicity. To facilitate this, we developed the Epitope Modification and MHC Prediction (EMMP) tool, which systematically evaluates amino acid substitutions for their predicted effects on major histocompatibility complex (MHC) binding. We applied this tool to modify a CD4⁺ T-cell epitope in the AAV9 capsid, which was identified and characterized as a proof-of-concept. Two mutant variants, R312H and R312Q, were selected and evaluated for transduction efficiency in vitro and immune response modulation in vivo. Notably, R312Q exhibited a significant reduction in T-cell activation and anti-AAV9 antibody production, albeit with a slight reduction in transduction at low multiplicities of infection (MOI). These results demonstrate a rational approach for optimizing AAV vector design, with potential applications for improvement the safety and efficacy of gene therapy.