Aim: Antibodies (Abs) against HLA-DQ are strongly associated with poor transplant outcomes. Eplets in the Epregistry are often referenced as the functional epitopes to describe these Abs; however, some specificity profiles derived from available single antigen bead (SAB) panels cannot be accounted for based on eplet analysis alone. Such deficits could be attributed to the gaps in the eplet system and/or the under-coverage of the existing SAB panels. Therefore, a new HLA-DQ SAB panel was designed to provide a proof-of-concept for functional epitope prediction.
Methods: Common HLA-DQ alpha and beta alleles are listed in CWD 2.0 and CIWD 3.0. The extracellular domains of all common HLA-DQ antigens (Ags) were subjected to a series of selection decisions using a proprietary algorithm to maximize the coverage of HLA-DQ amino acid polymorphism while minimizing the number of Ags. Selected Ags that could be recombinantly expressed on cell surface were included in a special build (SB) following standard LABScreenTM production protocols. Luminex assay was performed on 12 HLA-specific human monoclonal antibodies (mAbs) developed by Leiden University Medical Center. HLAMatchmaker and AminoAcid modules in HLA Fusion software were used to predict potential functional epitopes. Cells expressing engineered variants (EVs) with amino acid mutations were compared with wild types in flow cytometry to identify critical residues interacting with mAbs.
Results: SB provides additional coverage of potential functional epitopes defined by a single or a cluster of residues in a 3-dimensional space. Some functional epitopes of HLA-specific mAbs could only be predicted when SB results were included because SB provided critical differentiating specificities. In some cases, imputation with AminoAcid analysis yielded more productive predictions than HLAMatchmaker. Subsequent complete or partial loss-of-function of EVs vs. wildtype in flow cytometry pointed to the critical residues involved in these interactions.
Conclusion: SB HLA-DQ panel expands the coverage of common allele polymorphism and provides differentiating specificity profiles. More comprehensive coverage enables more relevant EV designs for the identification of critical residues recognized by mAbs. The proposed binding sites of the mAbs in this study should be confirmed by cryo-EM or crystallography.
