Central to the idea of antibody recognition is some degree of foreignness of the target antigen compared to the antibody producer. Epitopes are distinct regions on an antigen to which antibody can be elicited and bound. However, for HLA antigens, there is no consensus definition of what represents the minimal functional immunogenic unit of dissimilarity. To assess this in an unbiased way, we developed a reverse engineering software strategy based on donor specific antibodies defined by single antigen beads and full length genomic high resolution HLA typing by NGS of recipients and donors (332 transplant pairs). Starting with the ATG of Exon 1 and moving stepwise one amino acid at a time for each of the following triplets, the algorithm compared every possible amino acid triplet of the recipient and donor for 11 loci (A, B, C, DRB1, DRB3, DRB4, DRB5, DQA1, DQB1, DPA1, DPB1). Results were agnostic with respect to HLA class, not restricted to just the mature protein, and not influenced by existing maps (e.g., IMGT, or epitope models). We also developed web-based functions in the 17th IHIWS database to collect the unbiased triplets so that we could group the transplant pairs with the same donor specific antibodies and find shared triplets within the groups as potential core or essential epitopes that trigger the antibody formation. Profiling the pairs where the same DSA was identified led to identification of discrete amino acid triplets shared among the pairs irrespective of HLA match. The potential epitopes were mapped onto the 3D protein structure for reference.
Keywords: Algorithm; Antibody; Epitope; Sequencing; Transplant.
Copyright © 2019. Published by Elsevier Inc.