Antibodies against HLA cross-reactivity groups: From single antigen bead assay to immunoinformatics interpretation of epitopes. (May 2021)

• Record Type: Journal Article
• Authors: Usureau, Cédric; Lefèvre, Edgar; Top, Isabelle; Nikolski, Macha; Varlet, Pauline; Choukroun, Gabriel; Desoutter, Judith; Guillaume, Nicolas
• Is Part Of: Molecular immunology. Volume 133(2021)
• Page Start: 154
• Page End: 162
• Publication Date: 2021-05
• Subjects: Ab antibody -- HLA human leukocyte antigen -- MFI mean fluorescence intensity -- CREG cross-reactive group; Major histocompatibility complex -- Transplantation -- Antibody -- Epitope; Immunochemistry -- Periodicals; Molecular biology -- Periodicals; Immunochemistry -- Periodicals; Allergy and Immunology -- Periodicals; Molecular Biology -- Periodicals; Immunochimie -- Périodiques; Biologie moléculaire -- Périodiques; Immunochemistry; Molecular biology; Periodicals; Electronic journals; 571.96
• DOI: 10.1016/j.molimm.2021.02.020
• Languages: English
• ISSNs: 0161-5890

Abstract

Highlights: Spearman's ρ correlation is used for understanding anti-HLA cross-reactivities. Spearman's ρ clearly identifies serological subgroups. The dispersion of the Spearman ρ values is explained by an eplet approach. The Spearman's ρ values are linked to the number of matched eplets. Abstract: Identification of anti-human leukocyte antigen (HLA) antibodies (Abs) is based on Luminex™ technology. We used bioinformatics to (i) study the correlations of mean fluorescence intensities (MFIs) for all the possible allele pairs, and (ii) determine the degree of epitope homology between HLA antigens. Using MFI data on anti-HLA Abs from 6000 Luminex™ assays, we provide an updated overview of class I and II HLA antigen cross-reactivity in which each node corresponded to an allele and each link corresponded to a strong correlation between two alleles (Spearman's ρ > 0.8). We compared these correlations with the serological groups and the results of an epitope analysis. The strongest correlations concerned allele-specific Abs directed against the same antigen. For the HLA-A locus, the highest values of Spearman's ρ reflected broad specificity. For the HLA-B locus, graphs defined the HLA-Bw4 public epitope, and correlations between HLA-A and -B alleles were only present for beads with the same Bw4 public epitope. For the HLA-C locus, we identified two groups that differed with regard to their KIR ligand subclassification. Lastly, the HLA-DRB1 subgroups were part of a network. In the epitope analysis, Spearman's ρ was related to the number of matched epitopes within pairs of alleles. The combination of Spearman's ρ with simple, undirected graphing constitutes an effective tool for understanding routinely encountered cross-reactivity profiles. Based on this model, we have implemented an online data visualization tool available at http://cusureau.pythonanywhere.com/.