Background: Immunopeptidomics is the unbiased mass spectrometry-based study of Human Leukocyte Antigen (HLA)-bound peptide antigens. These peptides are presented for immune surveillance by T cells. Mutated or self-antigens bearing aberrant post-translational modifications (PTM), can elicit anti-tumour immune responses, particularly when associated with cancer-specific mutations or aberrant cancer-associated signaling events. However, analyzing PTM peptides, especially immunopeptides that have diverse termini (non-tryptic), is challenging. The traditional workflows in tandem mass spectrometry struggle with the loss of the PTM during fragmentation and, consequently, loss of accurate identification and localization. Therefore, this study leverages an alternative fragmentation method called Electron-Activated Dissociation (EAD) on the ZenoTOF 7600 system to assess its usability for PTM and non-PTM immunopeptides.
Methods: In our preliminary assessments, we used HLA-A*2:01 peptides isolated from a B-cell lymphoma to test the EAD reaction time and aimed to see whether combining EAD and Collision-Induced Dissociation (CID) would benefit over the use of EAD or CID fragmentation alone. In the present study, we analyzed HLA class I (HLA-B*57:01) and HLA class II (an HLA-DR, -DP, and -DQ mixture) peptides to assess a combination of kinetic energy (3eV vs. 6eV vs. 9eV) and the EAD reaction time (10ms vs. 20ms). Finally, we enriched HLA class II phosphopeptides using the Zr-IMAC system to assess the applicability of EAD-CID fragmentation further.
Results: In Class I peptides, we observe an EAD-assisted boost in the ratio of detected phosphorylated and cysteinylated peptides. Moreover, the combination of data derived from standard CID fragmentation with a second run using EAD-CID significantly boosted the number of confident peptide identifications. Application of hybrid EAD-CID fragmentation method to phospho-enriched immunopeptides demonstrated its potential for improvement of site localization and for teasing apart closely eluting phospho-isomers. Namely, we manually investigated a Class II phosphopeptide mapping to the C-terminus of KRTCAP2 with 4 possible modification sites. In this case, we observed that only two of the four possible sites (pT9 and pS17) have irrefutable spectral evidence. Of note, one of the two isomers (pS17) has not been reported previously in databases like PhosphoSitePlus.
Conclusion: we demonstrated that the hybrid EAD-CID fragmentation is versatile and tunable, and it improves immunopeptidome coverage when used in tandem with conventional LC-MS/MS workflows, and PTM site-localization, which is critical for the confident identification of novel PTM immunopeptides. These methods can be extended to clinical human tissues to investigate novel post-translationally modified immunopeptides in health and diseases like autoimmunity and cancer.