Computational approach for the design of a less immunogenic GALNS enzyme

Mucopolysaccharidosis IVA (MPS IVA) is a lysosomal disease (LSD) caused by the deficiency of the N-acetylgalactosamine-6-sulfate sulfatase (GALNS). Enzyme replacement therapy (ERT) represents the closest alternative for the specific treatment of this disease. Nevertheless, immune response to ERT limits the effectiveness of the treatment and might increase the cost of the therapy. We present a computational approach for the identification of human GALNS antibody epitopes and the in-silico modification as a strategy to decrease its immunogenicity. Epitopes were predicted using Epitopia Server and ElliPro tools. Amino acids that may reduce antigenicity were assessed for impact on protein structure by using several bioinformatics tools. Eleven antibody epitopes were identified in GALNS, with the C-terminal domain showing the highest antigenicity. The most immunogenic sequences were found in the protein surface; four of them are present in random-coils loops, while the remaining sequences were in α-helices and β-sheets regions. Only two amino acid changes allowed a reduction of antigenicity in the two most immunogenic epitopes without dramatically altering the structure of the peptide. Modeling of these amino acid changes within GALNS 3D model did not have significant impact in protein structure, with a RMSD of 0.03 Å against the unmodified enzyme. These results could have a significant impact in the design of less immunogenic recombinant enzymes for the development of an alternative therapeutics for MPS IVA and other LSD.