β-cell-specific CD8 T cell phenotype in type 1 diabetes reflects chronic autoantigen exposure

Abstract

Autoreactive CD8 T cells play a central role in the destruction of pancreatic islet β-cells that leads to type 1 diabetes, yet the key features of this immune-mediated process remain poorly defined. In this study, we combined high-definition polychromatic flow cytometry with ultrasensitive peptide-human leukocyte antigen class I tetramer staining to quantify and characterize β-cell-specific CD8 T cell populations in patients with recent-onset type 1 diabetes and healthy control subjects. Remarkably, we found that β-cell-specific CD8 T cell frequencies in peripheral blood were similar between subject groups. In contrast to healthy control subjects, however, patients with newly diagnosed type 1 diabetes displayed hallmarks of antigen-driven expansion uniquely within the β-cell-specific CD8 T cell compartment. Molecular analysis of selected β-cell-specific CD8 T cell populations further revealed highly skewed oligoclonal T cell receptor repertoires comprising exclusively private clonotypes. Collectively, these data identify novel and distinctive features of disease-relevant CD8 T cells that inform the immunopathogenesis of type 1 diabetes.

Figure 1

Identification of antigen-specific CD8 T cell populations. A-G: Thawed PBMCs were stained with pHLA-A*0201 tetramers representing PPI_15-24 (A), InsB_10-18 (B), IGRP_265-273 (C), IA-2_797-805 (D), GAD65_114-123 (E), CMV pp65_495-503 (F) and EBV BMLF1_280-288 (G). Gates were set serially on singlets, live CD3⁺CD14⁻CD19⁻ cells and lymphocytes prior to Boolean exclusion of dye aggregates and subsequent analysis in bivariate CD8 versus tetramer plots (Supplementary Figure 1 A). Representative paired data in the absence or presence of dasatinib are shown for type 1 diabetes patients and healthy controls. Tetramer-binding CD8 T cell frequencies are indicated.

Figure 2

Summary of antigen-specific CD8 T cell frequencies. A-G: Thawed PBMCs were stained with pHLA-A*0201 tetramers representing PPI_15-24 (A), InsB_10-18 (B), IGRP_265-273 (C), IA-2_797-805 (D), GAD65_114-123 (E), CMV pp65_495-503 (F) and EBV BMLF1_280-288 (G). Graphs show tetramer-binding CD8 T cell frequencies in the absence (empty symbols) or presence (filled symbols) of dasatinib for type 1 diabetes patients (T1D) and healthy controls (CTR). No significant differences across identical comparisons were observed between subject groups. Bars represent median values. Statistical analyses were performed using the Wilcoxon signed-rank test; p values ≤ 0.05 are shown.

Figure 3

Phenotypic subset analysis of antigen-specific CD8 T cells. A: Pie chart representations of mean subset percentages for pooled β-cell-specific CD8 T cells and individual specificities in type 1 diabetes patients (top row) and healthy controls (bottom row). Subsets are defined and color-coded as follows: green, T_N (CD27⁺CD45RO⁻CD57⁻CD95⁻CCR7⁺); yellow, T_SCM (CD27⁺CD45RO⁻CD95⁺CCR7⁺); blue, T_EFF (CD27⁻CD45RO⁻CD95⁺CCR7⁻); red, all remaining memory cells. B: Column plots showing T_N and T_SCM subset percentages for pooled β-cell-specific CD8 T cells and individual specificities in type 1 diabetes patients (T1D) and healthy controls (CTR). Representative data are shown in Supplementary Figure 1 B&C. Almost identical results were obtained with the T_N subset defined in the absence of CD57 (CD27⁺CD45RO⁻CD95⁻CCR7⁺). Bars represent median values. Statistical analyses were performed using the Mann-Whitney U-test; p values ≤ 0.05 are shown.

Figure 4

Single marker analysis of antigen-specific CD8 T cells. A-E: Percent expression of CD95 (A), CD57 (B), CD27 (C), CCR7 (D) and CD45RO (E) is shown for type 1 diabetes patients (T1D) and healthy controls (CTR) across all CD8 T cells and the indicated specificities. Corresponding data for individual β-cell-derived epitope-specific CD8 T cell populations are shown in Supplementary Figure 4. Bars represent median values. Statistical analyses were performed using the Mann-Whitney U-test; p values ≤ 0.05 are shown.

Figure 5

Frequency difference gating analysis of antigen-specific CD8 T cells. Overlays of concatenated data from type 1 diabetes patients (cloud plots) and healthy controls (blue dots) are shown for the indicated β-cell specificities across bivariate phenotypic profiles. Populations of CD27^intermediateCD95⁺ CD8 T cells, present at significantly higher frequencies in type 1 diabetes patients compared to healthy controls, are displayed as red dots.