Defects in IL-2R signaling contribute to diminished maintenance of FOXP3 expression in CD4(+)CD25(+) regulatory T-cells of type 1 diabetic subjects

Abstract

Objective: In humans, multiple genes in the interleukin (IL)-2/IL-2 receptor (IL-2R) pathway are associated with type 1 diabetes. However, no link between IL-2 responsiveness and CD4(+)CD25(+)FOXP3(+) regulatory T-cells (Tregs) has been demonstrated in type 1 diabetic subjects despite the role of these IL-2-dependent cells in controlling autoimmunity. Here, we address whether altered IL-2 responsiveness impacts persistence of FOXP3 expression in Tregs of type 1 diabetic subjects.

Research design and methods: Persistence of Tregs was assessed by culturing sorted CD4(+)CD25(hi) natural Tregs with IL-2 and measuring FOXP3 expression over time by flow cytometry for control and type 1 diabetic populations. The effects of IL-2 on FOXP3 induction were assessed 48 h after activation of CD4(+)CD25(-) T-cells with anti-CD3 antibody. Cytokine receptor expression and signaling upon exposure to IL-2, IL-7, and IL-15 were determined by flow cytometry and Western blot analysis.

Results: Maintenance of FOXP3 expression in CD4(+)CD25(+) Tregs of type 1 diabetic subjects was diminished in the presence of IL-2, but not IL-7. Impaired responsiveness was not linked to altered expression of the IL-2R complex. Instead, IL-2R signaling was reduced in Tregs and total CD4(+) T-cells of type 1 diabetic subjects. In some individuals, decreased signal transducer and activator of transcription 5 phosphorylation correlated with significantly higher expression of protein tyrosine phosphatase N2, a negative regulator of IL-2R signaling.

Conclusions: Aberrant IL-2R signaling in CD4(+) T-cells of type 1 diabetic subjects contributes to decreased persistence of FOXP3 expression that may impact establishment of tolerance. These findings suggest novel targets for treatment of type 1 diabetes within the IL-2R pathway and suggest that an altered IL-2R signaling signature may be a biomarker for type 1 diabetes.

FIG. 1.

Maintenance of FOXP3 expression in CD4⁺CD25⁺ nTregs of type 1 diabetic subjects is impaired in the presence of IL-2. As diagramed in (A), fresh CD4⁺CD25⁺ T-cells were sorted from PBMCs isolated from control and type 1 diabetic subjects, placed in culture with media alone or 200 IU/ml IL-2, and FOXP3 and CD25 expression was assessed over time by flow cytometry. B: One control and one type 1 diabetic sample are shown. C: CD4⁺CD25⁺ T-cells from multiple control (n = 17) and type 1 diabetic (n = 13) subjects were assayed as in A in the presence of media alone, 200 IU/ml IL-2, or 10 ng/ml IL-7. Bars show means and symbols represent individual subjects. Analysis was performed by gating on live CD4⁺CD25⁺ T-cells. Statistical significance was determined using an independent Student t test. Cohorts of control and type 1 diabetic subjects had mean ages of 38 (range 18–61) and 34 (range 21–46) years, respectively.

FIG. 2.

nTregs of type 1 diabetic subjects display impaired IL-2R signaling. A: Level of expression of CD25, CD122, and CD132 in the CD4⁺CD25⁺ population of control (n = 17) and type 1 diabetic (n = 11) subjects from Fig. 1C was determined using Quantum R-PE MESF beads. B: Thawed PBMCs from a representative control and type 1 diabetic subject were stimulated with 100 IU/ml IL-2 for 20 min prior to fixation and staining with CD4, CD25, and pSTAT5(Y694). Dashed lines are treatment with media alone and solid lines are treatment with IL-2. C: Multiple control (n = 15) and type 1 diabetic (n = 17) subjects were stimulated as in B. Bars show means and symbols represent individual subjects. Analysis of pSTAT5 was performed by gating on live CD4⁺CD25⁺ T-cells. Statistical significance was determined using an independent Student t test.

FIG. 3.

FOXP3 expression in iTregs of type 1 diabetic subjects is impaired in the presence of IL-2 but not IL-7. A: CD4⁺CD25⁻ T-cells were isolated from previously frozen control (n = 15) and type 1 diabetic (n = 13) subjects and activated with 5 μg/ml anti-CD3 antibody and irradiated accessory cells in the presence of media alone, 100 IU/ml IL-2, or 10 ng/ml IL-7. FOXP3 expression 48 h after activation was determined by flow cytometry by gating on live, total CD4⁺ T-cells. Asterisk denotes significant difference from media alone using a paired Student t test. Error bars represent means ± SEM. B: CD69 expression 48 h after activation was assessed by flow cytometry for a subset of control (n = 9) and type 1 diabetic (n = 10) subjects shown in A. C: Linear regression was performed for samples in B activated in the presence of 100 IU/ml IL-2 to determine the association between CD69 and FOXP3 expression for control (R² = 0.8496, P = 0.0004) and type 1 diabetic (R² = 0.1702, P = ns) subjects. The difference between the slopes of the lines was measured using an ANCOVA with the P value noted in the graph.

FIG. 4.

CD4⁺ T-cells of type 1 diabetic subjects display diminished responsiveness to IL-2. Thawed PBMCs from control subjects were stimulated with 100 IU/ml IL-2 for 10 min prior to fixation and staining for CD4, CD25, and pSTAT5(Y694). Analysis was performed by gating on total live CD4⁺ T-cells and comparing response to media alone versus cytokine stimulation. A: Staining for one representative sample is shown. B: The frequency of CD4⁺ T-cells that were pSTAT5(Y694)⁺ in response to IL-2 was determined for control (n = 59) and type 1 diabetic (n = 33) subjects. Bars represent means and symbols represent individual subjects. C: Control (n = 12) and type 1 diabetic (n = 13) subjects were assayed for pSTAT5(Y694) in response to stimulation for 10 min with 200 pg/ml IL-15 or 40 pg/ml IL-7. Bars represent means ± SEM. D: MFI fold increase of pSTAT5(Y694) in STAT5⁺ CD4⁺ T-cells of thawed PBMCs from control (n = 12) and type 1 diabetic (n = 14) subjects was determined by comparing pSTAT5(Y694) MFI after stimulation with IL-2, IL-7, and IL-15, or media alone. Bars show means and symbols represent individual subjects. All P values were determined using an independent Student t test.

FIG. 5.

Diminished IL-2 responsiveness is a stable phenotype and intrinsic property of CD4⁺ T-cells of type 1 diabetes. A: Thawed PBMCs from control (n = 18) and type 1 diabetic (n = 20) subjects were stimulated with 100 IU/ml IL-2 or 200 pg/ml IL-15 for 10 min prior to fixation and staining for CD4, CD25, and pSTAT5(Y694). Linear regression was performed to determine the relationship between pSTAT5 responses in the same cells after stimulation with either IL-2 or IL-15. Trend lines represent linear regression of control (R² = 0.3713, P = 0.0073) and type 1 diabetes (R² = 0.2003, P = 0.0478) data. B: Cells isolated from the same individual but at different dates were assayed as described (in A) for response to IL-2. Sample collection dates varied from 5 months to 6 years. The number of months between sample collection did not correlate with SD or coefficient of variation for the control subjects, type 1 diabetic subjects, or all samples combined as analyzed by linear regression.

FIG. 6.

Altered expression of molecules in the IL-2R signaling cascade in CD4⁺ T-cells of type 1 diabetic subjects. CD4⁺ T-cells were isolated from fresh PBMCs of control and type 1 diabetic subjects, and whole-cell protein lysates were analyzed by Western blot. Immunoblots were probed with JAK1-, JAK3-, and PTPN2-specific antibodies and an anti-TFIIB antibody as a loading control. Protein expression was determined by densitometry, normalizing each sample to TFIIB and expressing total protein levels relative to a Jurkat control present on each blot. Total JAK1 (A) and JAK3 (B) protein expression was compared between control (n = 20) and type 1 diabetic (n = 18 and 20 for JAK1 and JAK3, respectively) subjects. C: PTPN2 protein expression was compared between control (n = 21) and type 1 diabetic (n = 18) subjects. Significance was determined using an independent Student t test. D: Thawed PBMCs from these same samples were assayed for pSTAT5 upon exposure with 100 IU/ml IL-2 for 10 min as in Fig. 4. Using linear regression, protein expression was compared with pSTAT5 for control (n = 13) and type 1 diabetic (n = 15) subjects. Correlation between PTPN2 protein expression in the total population (control and type 1 diabetic subjects combined) and pSTAT5 is noted in the graph. Solid trend line (R² = 0.02, P = 0.6) and squares denote control subjects, and dashed trend line (R² = 0.234, P = 0.067) and open circles denote type 1 diabetic subjects. E: Thawed PBMCs from the type 1 diabetic subject with the highest PTPN2 expression or (F) type 1 diabetic subject with PTPN2 expression above the mean 0.8 (n = 6) were incubated with a PTPN2 inhibitor (compound 8 in [29]) for 30 min prior to stimulation with IL-2 for 10 min as in Fig. 4.