CD4 T Cells Reactive to Hybrid Insulin Peptides Are Indicators of Disease Activity in the NOD Mouse

Abstract

We recently established that hybrid insulin peptides (HIPs), formed in islet β-cells by fusion of insulin C-peptide fragments to peptides of chromogranin A or islet amyloid polypeptide, are ligands for diabetogenic CD4 T-cell clones. The goal of this study was to investigate whether HIP-reactive T cells were indicative of ongoing autoimmunity. MHC class II tetramers were used to investigate the presence, phenotype, and function of HIP-reactive and insulin-reactive T cells in NOD mice. Insulin-reactive T cells encounter their antigen early in disease, but they express FoxP3 and therefore may contribute to immune regulation. In contrast, HIP-reactive T cells are proinflammatory and highly diabetogenic in an adoptive transfer model. Because the frequency of antigen-experienced HIP-reactive T cells increases over progression of disease, they may serve as biomarkers of autoimmune diabetes.

Figure 1

HIP tet⁺ cells are present in the islets of NOD mice. A and C: Single-cell suspensions were prepared from the islets of nondiabetic NOD mice and stained with antibodies and tetramers. Tetramers contained the following peptides: HEL (HEL_11–25), 2.5HIP (LQTLALWSRMD), 6.9HIP (LQTLALNAARDP), and ins B:9-23 mimotopes (insp8G [HLVERLYLVCGGEG] and insp8E [HLVERLYLVCGEEG]). Gates were set on live, CD45⁺, lineage-negative (Lin⁻), CD4⁺ cells, as shown in Supplementary Fig. 1. A: Representative example of tetramer staining from the islets of one NOD mouse. Data are representative of five independent experiments. B: Single-cell suspensions were prepared from whole pancreas of young (3-, 6-, and 10-week-old) or diabetic NOD mice and stained with antibodies and tetramers. Each symbol represents an individual mouse, and a summary of six independent experiments is provided. C: Tetramer analysis in the islets of 17 nondiabetic NOD mice (aged 6–20 weeks); data are a summary of four independent experiments. *P < 0.05; ***P < 0.001.

Figure 2

Phenotype of tet⁺ cells in the pLN vs. bLN at 3, 6, and 10 weeks in NOD mice. A: Single-cell suspensions were prepared from pLN (panc LN) or bLN (brac LN) from a 10-week-old NOD mouse. Cells were stained with antibodies and tetramers and then analyzed by flow cytometry. Tetramers contained the following peptides: 2.5HIP (LQTLALWSRMD), 6.9HIP (LQTLALNAARDP), and ins B:9-23 mimotopes (insp8G [HLVERLYLVCGGEG] and insp8E [HLVERLYLVCGEEG]). Gates were set on live, CD45⁺, lineage-negative (Lin⁻), CD4⁺ cells. The percentage of CD44^hiCD62L^lo is reported for each population (numbers within boxes). B: Summary of three independent experiments per age-group for 3-week-old (n = 9), 6-week-old (n = 12), and 10-week-old (n = 11) NOD mice. Each symbol represents an individual mouse. ***P < 0.001.

Figure 3

HIP-reactive T cells are devoid of Tregs and secrete IFN-γ. A: FoxP3 expression in the spleen and pancreas (Pancr) of NOD or NOD.ChgA^−/− mice from a representative experiment. Gates were set on live, CD45⁺, lineage-negative (Lin⁻), CD4⁺ cells. B: FoxP3 expression in the pancreas and spleen in nondiabetic (12- to 20-week-old) NOD mice. Data are a summary of at least three independent experiments. Tetramers contained the following peptides: 2.5HIP (LQTLALWSRMD), 6.9HIP (LQTLALNAARDP), or ins B:9-23 mimotopes (insp8G [HLVERLYLVCGGEG] and insp8E [HLVERLYLVCGEEG]). C: FoxP3 expression in ChgA-sufficient or ChgA knockout (ChgAKO) mice. Experiments were performed using diabetic wild-type NOD mice or aged-matched NOD.ChgA^−/− mice. D–F: Splenocytes from diabetic NOD mice were cultured with or without antigen (Ag) in an IFN-γ ELISpot assay and the number of spots (cytokine-secreting cells) are reported. D: Assay plate wells from a representative experiment. Numbers to the upper right of each well indicate the number of spots counted. Duplicate wells are shown. E: Summary of three independent experiments. n = 10 mice. F: Correlation between the number of 2.5HIP tet⁺ events obtained by flow cytometry vs. number of spots obtained by ELISpot. Data are expressed per 10⁶ splenocytes. *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 4

Antigen specificity and diabetogenicity of HIP-reactive T cells. A: tet⁺ cells are present in three BDC T-cell lines. Single-cell suspensions were stained with I-A^g7 tetramer. After 45 min, cells were counterstained with anti-CD4 and a viability dye before flow cytometry analysis. Gates were set on live CD4⁺ cells, and data are representative of two independent experiments for each T-cell line. Our results indicate that all three lines contained 2.5HIP tet⁺ cells, and both BDC-9 and BDC-10 contained 6.9HIP tet⁺ cells. tet⁺ cells were then sorted by FACS to obtain pure populations of tet⁺ cells. B: The islet-reactive T-cell clones PD-12.4.4 (insulin-reactive), BDC-6.9 (6.9HIP-reactive), and BDC-9.H1 were stained with antibodies and tetramers. Gates were set on live CD4⁺ cells, and data are representative of three independent experiments for each T-cell clone. C: To verify clonality, BDC-6.9 and BDC-9.H1 were stained with Vβ4 or Vβ6 antibodies. Gates were set on live CD4⁺ cells, and data are representative of at least four independent experiments for each T-cell clone. D: BDC-4.38 (insulin-reactive), BDC-6.9, or BDC-9.H1 T-cell clones from the BDC panel were challenged with NOD or NOD.IAPP^−/− islets, and IFN-γ secretion was measured by ELISA. Data are representative of three independent experiments. E: The T-cell clone BDC-9.H1 was challenged with peptide antigens, and IFN-γ was measured by ELISA. Peptides used were insulin C-peptide and IAPP2 (added together in the same culture well), the 6.9 HIP core peptide (LQTLALNAARDP), or the full-length 6.9HIP (EVEDPQVAQLELGGGPGAGDLQTLALNAARDPNRESLDFLLV). Data are representative of three independent experiments. F: The T-cell clone BDC-11.1 is 2.5HIP reactive. The T-cell clone BDC-11.1 (2 × 10⁴) was challenged with antigen (Ag) and antigen-presenting cells (2 × 10⁴). After 24 h, supernatants from cell cultures were tested for the presence of IFN-γ by ELISA. Results are from duplicate wells from two independent experiments. G: The T-cell clones BDC-9.H1 or BDC-11.1 were adoptively transferred into young (6–14 days old) NOD mice (n = 3 recipient mice/group from two independent experiments per T-cell clone), and mice were monitored daily for signs of hyperglycemia. Mice were considered diabetic when blood glucose levels rose above 15 mmol/L. KO, knockout.

Figure 5

Frequency and phenotype of tet⁺ cells in the blood of NOD mice over time. Peripheral blood from NOD mice was collected by submandibular bleed, and PBMCs were isolated on a Ficoll gradient. PBMCs were then stained with antibodies and tetramers. Tetramers contained the following peptides: 2.5HIP (LQTLALWSRMD), 6.9HIP (LQTLALNAARDP), and ins B:9-23 mimotopes (insp8G [HLVERLYLVCGEEG] and insp8E [HLVERLYLVCGEEG]). A: A representative example of the frequency of tet⁺ cells in the PBMCs from a mouse in the cohort. B: A representative example of CD44 vs. CD62L expressed on mouse PBMCs. Gates were set on live, CD45⁺, CD4⁺, lineage-negative (Lin⁻) cells or on tet⁺ cells (as indicated). C: Disease incidence in the cohort of 14 mice is shown in the left panel. The percentage of tet⁺ cells is shown in the center panel, and the percentage of CD44^hiCD62L^lo cells is represented in the right panel. Asterisks indicate significant differences compared with insulin tet⁺ cells (P < 0.05). *P < 0.05.