Re-definition of the DR3, DR4, DR9 and DR11 motifs by cluster analysis allows the identification of the hidden DR52 and DR53 peptide repertoires

Abstract

HLA-DR alleles are commonly inherited as haplotypes. While HLA-DR3 (DRB1*0301) and -DR11 (DRB1*1101) are linked with different alleles of HLA-DR52 (DRB3), HLA-DR4 (DRB1*0401) and -DR9 (DRB1*0901) are linked with HLA-DR53 (DRB4). The overexpression of the DRB1-encoded molecules on the cell surface has hindered the definition of the peptide binding motifs for the non-DRB1 molecules because of the difficulty to distinguish the peptides corresponding to each isotype. The definition of these hidden motifs is of vital importance because of the recently description in bibliography of cancer or autoimmunity related peptides supposedly presented by these non-DRB1 molecules. Here, we show an innovative bioinformatics clustering analysis that allows such discrimination. We isolated class II-peptide complexes from HLA-DR homozygous B cell lines by immunoprecipitation with an HLA-DR-specific antibody and we identified the binding peptides by mass spectrometry. Later, the eluted peptides were analysed with the GibbsCluster-2.0 software. This workflow let us to differentiate peptides that belong to DR3 or DR11 from those of the DR52 isotype, as well as the ones from DR4 or DR9 and DR53. Our results confirmed the DR3 and DR4 motifs and improved the accuracy of the, since this moment, incompletely defined DR9 and DR11 motifs. Moreover, we were able to establish the motifs corresponding to HLA-DR52 alleles, in addition to HLA-DR53. Surprisingly, we have found a distribution of the conserved positions in the non-DRB1 motifs (P1, P7 and P9) different from the typically found in DRB1 molecules (P1, P4/P6 and P9), highlighting a proline and an aspartic acid in P7 of the HLA-DR52 and HLA-DR53 molecules, respectively. This phenomenon is absolutely uncommon in DRB1 alleles. In conclusion, we have developed a new strategy to define hidden HLA-DR motifs that could enable to show up a novel landscape of target peptides capable of being used in immunotherapy.