Session 11: Abstract presentations

CD8+T cells survey the repertoire of antigenic peptides on beta cells presented in the context of HLA-I. During insulitis, metabolic or inflammatory stress in the vicinity of beta cells induces local enzymatic activities and promotes the generation of post-translationally modified (PTM) neoantigens to which central immune tolerance is absent. PTM of antigens is known to increase the diversity of peptides presented in the periphery and as such can contribute to the loss of T cell tolerance, culminating in autoimmunity. Importantly, the natural repertoire of islet cell presented antigenic peptides has not been studied. Here, we used a reverse immunology strategy in combination with highly sensitive mass spectrometry to characterise the HLA-A*02:01 immunopeptidome of cytokine-conditioned human islets. We identified a PTM PPI15-24 epitope that is naturally processed and presented by HLA-A*02:01 on human islet cells. The PTM PPI15-24 epitope is generated by an islet cell resident enzyme, the activity of which can be induced by inflammatory stress. Intriguingly, despite very close structural similarity between A2/PPI15-24 and A2/(PPI15-24)PTM, the highly promiscuous PPI15-24-specific 1E6 CD8+ T cell clone does not recognise the PTM-PPI15-24. However, tetramer staining experiments show that PTM-PPI15-24 is a major target for circulating effector CD8+ T cells from HLA-A*02:01 patients with new onset of type 1 diabetes, indicating that it is a target for pathogenic autoreactivity. Work is ongoing to isolate and characterise patient derived CD8 T cells specific for PTM-PPI15-24. This study highlights the fact that beta cell specific PTMs generated by an inflammation inducible enzyme can be targeted by autoreactive T cells clonally distinct from those targeting wild type epitopes. A better understanding of these T cells will aid in biomarker development and potentially in the development of therapeutic interventions.

Background: Most pathogenic T cells in NOD mice recognize an insulin B:9-23 derived peptide bound to the I-Ag7 MHC class II molecule in the weak affinity register 3 (R3). The IAg7/B:9-23(R3) complex contains two epitopes based on peptides with or without the native glutamate at position B21 or the core peptides termed the P8E and P8G epitopes. Previously we demonstrated that an I-Ag7/B:9-23(R3) specific monoclonal antibody mAb287 could partially protect recipient NOD mice from T1D. Our later structural analyses indicate that mAb287 only binds to p8E epitope but not the p8G epitope. Thus, we investigated if advanced antibody-based therapies with improved features are able to protect disease efficiently.
Methods: To increase therapeutic efficacy we have explored two alternative approaches. First, we have transferred redirected CD8 T cells expressing a mAb287 chimeric antigen receptor (287-CAR) to pre-diabetic mice. Second, we have treated mice with another inhibitory antibody, mAb7E, that can bind both p8E and p8G epitopes.
Results: 1) 287-CAR T cells maintain the binding specificity of mAb287, and selectively lyse antigen presenting cells expressing I-Ag7/B:9-23(R3) in vitro. In vivo, 287-CAR-T cells travel to pancreatic lymph nodes, and a single infusion of CAR-T cells can significantly delay, although not prevent, the onset of T1D. 2) MAb7E has higher binding affinities with both p8E and P8G epitopes than mAb287 and suppresses both p8E and p8G reactive T cells in vitro. In vivo, mAb7E provides greater protection from T1D than mAb287 (70% vs 35%) when administered beginning either early (4 weeks) or late (9 weeks) during pre-diabetes stages. 
Conclusion: Antibody-based immune interventions targeting I-Ag7/B:9-23(R3), with both the bi-specific antibody and antibody redirected T cells, are able to modulate the development of spontaneous T1D safely. Current studies are directed towards testing reagents recognizing orthologous HLA-DQ8/insulin complexes for the future clinical application.

Background - The antigens and epitopes recognized by CD4+ T cells in people with type 1 diabetes (T1D) remain poorly defined. We, and others, have found that several epitopes derived from C-peptide are recognized by human islet-infiltrating CD4+ T cells implicating C-peptide as an autoantigen in human T1D. Here we quantified and analyzed CD4+ T cells that responded to full-length C-peptide (PI33-63) in peripheral blood of subjects with and without T1D.
Methods - The CFSE dye-based proliferation assay was used to detect C-peptide-responsive CD4+ T cells in PBMC (1). CD4+ T cells that responded to C-peptide were sorted into individual wells, cloned and characterised.
Results - CD4+ T-cell responses to full-length C-peptide were detected in: >60% (14 of 23) of recent-onset (>100 days of diagnosis) T1D subjects, 13% (2 of 15) of long-standing (>100 days from diagnosis) T1D subjects and 8% (1 of 13) of HLA-matched healthy subjects. A panel of 22 C-peptide-specific CD4+ T-cell clones were generated from 6 individuals with recent-onset T1D. These clones recognized epitopes across the entire 31 amino acids of C-peptide, although most epitopes were towards the C-terminus. Eighty-six percent (19 of 22) C-peptide-specific clones were restricted by high-risk alleles HLA-DQ8, -DQ2, -DQ8trans or -DQ2trans. TCR sequencing revealed that these clones used a wide variety of TCR genes. Finally, titration experiments showed that full-length C-peptide was a much more potent agonist of some CD4+ T-cell clones than an 18mer peptide encompassing their cognate epitope.
Conclusion  - Our findings support the notion that full-length C-peptide is an important target of pathogenic CD4+ T cells in people with DQ8 and/or DQ2 who develop T1D. Consequently, full-length C-peptide may be useful in T-cell assays and antigen-specific therapy protocols.
1. S. I. Mannering, et al. A sensitive method for detecting proliferation of rare autoantigen-specific human T cells. J Immunol Methods 283, 173-183 (2003).

Immune therapies for type 1 diabetes (T1D) are being tested for their ability to preserve pancreatic beta-cell function. Beta-cell function is most commonly measured using a meal test to determine the plasma C-peptide area under the curve (CPAUC). This requires 7 blood samples and takes 3 hours to perform. A simpler measure of beta-cell function is needed to facilitate implementation of effective immune therapies in the clinic. To determine if CPAUC could be reliably estimated using measurements from a single fasting clinical assessment, we developed and validated linear models using routine biochemical and clinical data from eight randomised therapy trials involving participants with recently-diagnosed T1D. A model based on fasting plasma C-peptide, fasting plasma glucose, HbA1c, disease duration, body mass index and insulin dose most accurately estimated loss of beta-cell function (area under ROC 0.89; 95% CI 0.87 to 0.92) and was superior to the reported ‘insulin dose-adjusted HbA1c’ (IDAA1C; area under ROC 0.72; 95% CI 0.68 to 0.76). When applied to data from randomized trials, model-estimated CPAUC (CPEST) reliably identified effective therapies and, using conservative estimates, required a 17% increase in sample size to meet statistical power requirements. Thus, CPEST approximated from six parameters at a single, fasting time point accurately identifies loss of beta-cell function and is comparable to observed CPAUC for identifying treatment effects. CPEST could therefore be used routinely to measure of beta-cell function in the clinic and serve as a simpler primary outcome measure for trials of disease-modifying therapy in T1D.