Session 2: Unconventional Antigens in Type 1 Diabetes

Hybrid insulin peptides (HIPs) are a new class of antigens that are targeted by autoreactive T cells isolated from the residual islets of organ donors with type 1 diabetes (T1D). Each HIP consists of a proinsulin fragment that is covalently cross-linked to another protein fragments through a peptide bond. We previously isolated and identified various HIPs in murine beta cell tumors and we demonstrated that several of these HIPs are targeted by a number diabetes-triggering CD4 T cell clones obtained from non-obese diabetic (NOD) mice. We also identified HIP-reactive T cells that were isolated from the residual islets of T1D organ donors, signifying localization of HIPs to the core of the infiltrated pancreatic islets in humans. We are currently applying mass spectrometric methods to determine if HIPs form in human islets. Additionally, we are using ELISPOT analyses to determine if new onset T1D patients have HIP-reactive T cells in their peripheral blood mononuclear cell (PBMCs). Identification of new HIPs in human islets and new HIP-reactive T cell specificities in T1D patients may provide us with reagents to further characterize the T cell mediated attack on beta cells and to test new strategies for antigen-specific tolerance induction.

Proteasome is the main producer of the epitopes presented on MHC-I molecules to the CD8+ T cells. The epitopes could be produced by peptide-bond hydrolysis or by proteasome-catalysed peptide splicing (PCPS). Proteasome-generated antigenic spliced peptides seem to be much more frequent than expected. Therefore, they can extend the antigenic landscape of cells to a grey area where self and non-self could be misjudged by the immune system. Do we have enough information to reckon if PCPS plays a relevant role in autoimmunity? Can the investigation of PCPS lead to the discovery of new autoantigens in the contest of T1D? Is there any relation between the Hybrid Insulin Peptides discovered by DeLong and colleagues, and presented onto MHC-II complexes, and proteasome-generated spliced peptides, which are presented by MHC-I complexes?

The cellular immune response relies upon T cell recognition of peptides presented on the cell surface in complex with HLA molecules. As such, it is the peptide cargo of these HLA molecules that dictates the quality of the immune response and ultimately the efficacy of protective immunity. Relatively simplistic models have been used to explain how these peptide antigens are generated and selected for presentation, however, such models fail to predict and explain the diversity and complexity of the immune response. Much of this unexplained complexity resides in degenerate and reconstructive proteolysis. We have recently highlighted the complex role of proteolysis in the generation and diversification of peptide antigens displayed for T cell recognition. For example, functional heterogeneity is observed during degenerate peptidase trimming of T cell epitopes such that peptides with ragged N- or C-termini (nested peptides) can be presented with different immune outcomes. Moreover, we have generated evidence that reinforces recent surprising studies that up to 30% of peptides presented by class I HLA molecules are generated by post-translational proteasomal splicing (i.e. the ligation of peptide fragments within the proteasome rather than peptide destruction). Using a novel bioinformatic workflow we show that these spliced peptides can be generated either from two regions of the same antigen (cis-splicing) or two distinct antigens (trans-splicing) by retrospectively interrogating datasets for several common HLA allotypes. This is the first demonstration that trans-spliced peptides are abundantly represented in the immunopeptidome. However, the role of spliced peptides in immune responses and the precise mechanism of their generation remains poorly understood and will be discuss in the context of Type 1 Diabetes and other immune disorders.