Preprint Club
A cross-institutional Journal Club Initiative

Keywords

• Type 1 diabetes

• Memory T cells

• Autoimmunity

• Diabetes neoantigen

Main Findings

Type 1 diabetes (T1DM) is a chronic condition caused by the progressive immune-mediated destruction of insulin-producing pancreatic β-islets. Although T cells are thought to significantly contribute to disease pathology, the nature of the immunostimulatory β-cell antigens, as well as how and where these antigens originate in the pancreas remains unclear. In this preprint, the authors applied a β-cell degranulation assay and HLA-II immunopeptidomics to identify potential islet-derived neonatigens that may be associated with the development of type 1 diabetes. To accomplish this, the authors recruited both non-diabetic (ND) individuals and diabetic patients who had been diagnosed either 3 or 18 months earlier. All participants underwent a mixed-meal tolerance test (MMTT), which involved fasting for 8 hours followed by consuming a carbohydrate-rich nutritional drink. The authors reasoned that the use of a MMTT would allow peptides from pancreatic β-cells to be released into the circulation, bind to HLA-II on circulating APCs. Through such a process the authors anticipated the discovery of pancreatic neoantigens via HLA-II-associated immunopepidomics. Curiously, along with native insulin B-chain peptides, the authors identified a previously unidentified modification of the insulin B-chain that was associated with HLA-II only following the MMTT, which contained a cysteine to serine transformation at the 19thamino acid (insulin C19S). Analysis of the murine immunopeptidomcis following glucose exposure revealed that the insulin C19S antigen is conserved across species and associated with mouse MHCII.

The authors then asked whether the insulin C19S peptide associates with disease progression in non-obese diabetic (NOD) mice, known for their spontaneous development of diabetes with age. To do this, the authors first generated C19S-specific T cell hybridomas (termed S5 T cells), which they verified to be specific to the C19S insulin peptide, but not native insulin. Then, pancreatic islet lysates from 5-week-old NOD mice (non-diabetic) and 17-week-old NOD mice (diabetic) were used to stimulate S5 T cells and assay the degree of T cell autoreactivity by IL-2 secretion. Interestingly, diabetic NOD mouse islets induced a greater degree of S5 T cell proliferation compared to non-diabetic NOD mice. Furthermore, antigenic ex vivo recall experiments and ELISPOTs analyses showed that C19S-specific T cells have a greater propensity to secrete IFN𝛾 relative to native insulin-specific T cells. Taken together, these data indicate that the C19S peptide is associated with diabetes disease progression.

In a separate set of experiments, the authors used the reactivity of S5 T cells to C19S to examine the precise location of the neoantigenic peptide within pancreatic β-cells, and the environmental conditions required for its synthesis. Indeed, the authors found that the C19S insulin neoantigen appears following a synergy of endoplasmic reticulum and oxidative stress that can occur in response to inflammatory cytokine signalling and primarily localizes within islet crinosomes.

The study then assessed the transcriptional and functional profiles of C19S-specific T cells to provide potential clues as to how this neoantigen may contribute to disease progression. A scRNA-sequencing based comparison of C19S-specific T cells to native insulin-specific T cells isolated from NOD mice lymph nodes indicated that the neoantigen specific T cells displayed higher transcriptional signatures of T cell activation and memory relative to native insulin-specific T cells.

Further experiments involving the adoptive transfer of C19S-specific T cells into NOD.Rag1-/-mice indicated that these cells are sufficient to induce diabetes in mice. Comparison of WT NOD and NOD.Tnfsf1a/1b-/- mice indicated that the heightened inflammatory activation and memory phenotype of C19S-specific T cells is dependent on inflammatory cytokine signalling.

The authors conclude their study by performing a series of impactful experiments using human patient samples. First, by utilizing human islets, crinosome fractionation, and crinosome peptidomics, the authors verify that the C19S peptide is localized within islet crinosomes in humans, and that its presence is induced following endoplasmic reticulum and oxidative stress, or following stimulation with pro-inflammatory cytokines such as TNF, IL-1β, and IFN𝛾.

Lastly, the authors examine the numbers and profile of C19S neoantigen-reactive T cells within PBMCs from non-diabetic, recent-onset, and established diabetes patient. Interestingly, both recent-onset and established diabetes populations have larger numbers of C19S reactive T cells than the non-diabetic population. Furthermore, in diabetes, there are more C19S-reactive T cells than there are native-insulin reactive T cells, implicating the presence of the neoantigen-specific cells as indicative of disease. Lastly, the C19S-specific T cells display a greater activated memory phenotype, and a lower propensity toward regulatory T cell phenotypes than native insulin-specific T cells in patients.

Limitations

The authors may consider commenting on why non-diabetic human islets produce the insulin C19S neonatigenic peptide, yet do not have as many C19S-specific T cells as the diabetic cohorts.

· The authors may consider examining whether the administration of antioxidants to NOD mice decreases the rate of disease development, or the ameliorates the pathology of disease.

One wonders whether the neoantigen or the degree of the β-cell associated stress are critical in causing the break in tolerance to insulin. Would over expression of the neoantigen be sufficient to trigger T1D. Delineating whether neoantigen or stress cooperate in T1D onset would be a major addition to this already significant finding.

Significance/Novelty

The preprint reveals that oxidative stress in pancreatic islets modifies insulin (C19S), creating a neoantigen that activates memory T cells in type 1 diabetes. Presented by HLA-DQ8, these cells persist, sustaining autoimmunity. Targeting oxidative antigen remodeling offers a novel strategy to modulate chronic immune responses in diabetes therapy.

This study shows how oxidative stress alters islet self-antigens, fuelling chronic autoimmunity. The study provides intriguing avenues toward targeting insulin neoantigen formation which could slow type 1 diabetes progression. This discovery shifts focus from immune suppression to precision strategies that block disease-driving antigen remodelling, offering new avenues for treatment and immune modulation.

Credit

Reviewed by Boyan Tsankov as part of a cross-institutional journal club between the Icahn School of Medicine at Mount Sinai, the University of Oxford, the Karolinska Institute the University of Toronto, and MD Anderson Cancer Centre.

The author declares no conflict of interests in relation to their involvement in the review.