Wan, X., Vomund, A.N., Peterson, O.J. et al. The MHC-II peptidome of pancreatic islets identifies key features of autoimmune peptides.Nat Immunol (2020). https://doi.org/10.1038/s41590-020-0623-7
In a publication from a recent edition of Nature Immunology, the authors used PEAKS X to investigate the peptidome of diabetes-associated class II major histocompatibility complex (MHC-II) alleles in pancreatic islets, lymph nodes, and spleens of non-obese diabetic mice. The MHC-II genetic variant, known as I-Ag7 in mice and HLA-DQ8 in humans, alters the binding specificity of tissue-derived peptides. The development of diabetes and autoimmune targeting of pancreatic β-cells is driven by CD4+ T cell recognition of insulin peptides displayed on antigen-presenting cells (APCs). Until now, exact MHC-II epitopes that trigger autoimmunity and development of diabetes remained unknown.
The authors discover an enrichment of β-cell derived peptides from the MHC-II molecules of APCs isolated from pancreatic islets. These peptides were largely from InsB and InsC, which are major forms of proinsulin. Varying lengths of epitopes from InsB and InsC were identified, however, CD4+ T cell reactivity was centered around InsB:12-20 and C-terminal Ins1C peptides. Importantly, the immunogenic insulin peptides were also present on APCs of peripheral lymphoid tissue, further contributing to T cell reactivity. Consistent with this, removal of peripheral lymph nodes decreases autoimmune diabetes development.
The authors used PEAKS PTM to discover a naturally occurring deamidation event that improves binding of Ins1C peptides to I-Ag7 MHC-II molecules. Moreover, a stringent multi-round search using PEAKS X validated the identity of hybrid insulin peptides and uncovered a few fused peptides in crinosomes and in the secretome. This was done by first searching mass spectra against the UniProt/SwissProt mouse proteome and to an in silico generated hybrid insulin peptide (HIP) database. This was followed by a search for alternative spectrum matches using PEAKS PTM and SPIDER algorithms. This method efficiently identifies unique sequence alternatives not found in the searched databases and was important for removing falsely assigned spectra. This article demonstrates the power of PEAKS X in accurately identifying novel immunogenic peptides with significant disease relevance.