Improved immunopeptidomics analysis methods for cancer antigen discovery: "What is your quest?" - "To seek the Holy Grail."

HLA peptide ligands which are presented abundantly, exclusively and uniformly on the tumor cell surface provide an essential starting point in the design of personalized immunotherapy. In this thesis, we have attempted to tackle several analytical challenges in the field of identifying HLA antigens, also termed immunopeptidomics. Based on these advances we have applied novel strategies to the translational context of colorectal cancer. In HLA peptide isolation and measurement techniques applied, we have incorporated high-pH fractionation, which is complementary to the widely used strong-cation exchange (SCX) approach. This allowed us to analyze the highly complex HLA peptide ligandome to greater depth. By using a pulldown flowthrough ‘recycling’ approach, in which HLA ligands are purified from the same sample in multiple rounds of purification, we could further maximize the amount of HLA peptide ligands isolated from a single sample. This may be applied critically to increase the yield of isolated HLA peptides for analysis, especially from clinical samples, where the amount of material is often a limiting factor. These developments, used in combination in a different study, allowed us to obtain new mechanistic insights into HLA loading in an artificial ‘fever-induced’ state. We first challenged a cell line system with high temperature to mimic a fever, and consequently detected specific changes in the HLA class II ligandome, mediated through alteration of the CD74 HLA class II chaperone. In this simple context of temperature effects on HLA peptide presentation, we could also focus on the presentation of peptides derived from HLA proteins themselves, taking on a frequently neglected angle than regular mass spectrometry-based ligandomics. By further applying glycoproteomics to characterize HLA protein molecules, we have also characterized their distinct glycan compositions, and shown that different HLA molecules have different glycosylation patterns, in strong correlation with their physical location inside the cell. Besides using mass spectrometry to understand the fundamental rules in HLA peptide processing and loading in different physiological states, as well as glycosylation-dependent translocation of peptide-bound HLA molecules, the tools established also made it possible to explore cancer antigen discovery in the patient context. In addition to the fundamental research questions, we also delved deep to look for clinically relevant tumor specific antigens and neo-antigens, using the organoid system as means to amplify limited patient material. Despite gains in ligand identification brought about by advances in instrumentation and sample preparation, the routine discovery of neo-antigens from patient derived biopsies remains challenging.