Assessing high affinity binding to HLA-DQ2.5 by a novel peptide library based approach

Here we report on a novel peptide library based method for HLA class II binding motif identification. The
approach is based on water soluble HLA class II molecules and soluble dedicated peptide libraries. A high
number of different synthetic peptides are competing to interact with a limited amount of HLA molecules,
giving a selective force in the binding. The peptide libraries can be designed so that the sequence
length, the alignment of binding registers, the numbers and composition of random positions are controlled,
and also modified amino acids can be included. Selected library peptides bound to HLA are then
isolated by size exclusion chromatography and sequenced by tandem mass spectrometry online coupled
to liquid chromatography. The MS/MS data are subsequently searched against a library defined database
using a search engine such as Mascot, followed by manual inspection of the results. We used two dodecamer
and two decamer peptide libraries and HLA-DQ2.5 to test possibilities and limits of this method.
The selected sequences which we identified in the fraction eluted from HLA-DQ2.5 showed a higher average
of their predicted binding affinity values compared to the original peptide library. The eluted
sequences fit very well with the previously described HLA-DQ2.5 peptide binding motif. This novel
method, limited by library complexity and sensitivity of mass spectrometry, allows the analysis of several
thousand synthetic sequences concomitantly in a simple water soluble format.