Design of new high-affinity peptide ligands for human leukocyte antigen-DQ2 using a positional scanning peptide library

Abstract

Human leukocyte antigen (HLA)-DQ2 (DQA1 x 0501/DQB1 x 0201) is associated with several immune disorders, including celiac disease, which is caused by an inappropriate T-cell response to gluten. Interference with peptide presentation by HLA-DQ2, for example, by the use of peptide blockers, is a possible treatment strategy for such HLA-associated disorders. A successful implementation of this strategy will depend on the identification of ligands that bind much better to HLA-DQ2 than the disease related epitopes. We have used a positional scanning nonapeptide library to determine the optimal amino acids for each position of the HLA-DQ2 binding frame. By combining the optimal residues in each position, we were able to design high affinity binders to HLA-DQ2. Interestingly, the decapeptide with highest affinity was composed of the most favorable residues in each position. This sequence bound 50-fold better than the immunodominant gluten epitope DQ2-alpha-I-gliadin, which makes it an interesting lead compound for the development of blockers. For some natural HLA-DQ2 ligands, the correlation between measured and predicted affinities was poorer, but notably these peptides did not have optimal amino acids at all positions. Our approach represents a straightforward strategy for developing high-affinity binders to HLA class II molecules.