Interleukin-35 administration counteracts established murine type 1 diabetes--possible involvement of regulatory T cells

Abstract

The anti-inflammatory cytokine IL-35 is produced by regulatory T (Treg) cells to suppress autoimmune and inflammatory responses. The role of IL-35 in type 1 diabetes (T1D) remains to be answered. To elucidate this, we investigated the kinetics of Treg cell response in the multiple low dose streptozotocin induced (MLDSTZ) T1D model and measured the levels of IL-35 in human T1D patients. We found that Treg cells were increased in MLDSTZ mice. However, the Treg cells showed a decreased production of anti-inflammatory (IL-10, IL-35, TGF-β) and increased pro-inflammatory (IFN-γ, IL-2, IL-17) cytokines, indicating a phenotypic shift of Treg cells under T1D condition. IL-35 administration effectively both prevented development of, and counteracted established MLDSTZ T1D, seemingly by induction of Eos expression and IL-35 production in Treg cells, thus reversing the phenotypic shift of the Treg cells. IL-35 administration reversed established hyperglycemia in NOD mouse model of T1D. Moreover, circulating IL-35 levels were decreased in human T1D patients compared to healthy controls. These findings suggest that insufficient IL-35 levels play a pivotal role in the development of T1D and that treatment with IL-35 should be investigated in treatment of T1D and other autoimmune diseases.

Figure 1. Treg cells are increased in MLDSTZ induced T1D.

(A) Male CD-1 mice were injected i.p. with STZ (40 mg/kg/day) or 200 μl saline, for 5 consecutive days. The blood glucose levels (mM) were monitored in vehicle (closed circles) or MLDSTZ (open circles) treated mice on days 0, 3, 7, 10, 14, 17 and 21 after the first injection of STZ. (B) Representative light micrograph of pancreatic islets from vehicle (left panel) or MLDSTZ (right panel) treated mice on day 21 (hematoxylin and eosin staining; magnification ×100). (C) The proportions of CD4⁺CD25⁺Foxp3⁺ Treg cells in thymocytes, PDLNs cells and splenocytes were analyzed by flow cytometry in vehicle or MLDSTZ treated mice on days 0–21. (D) Real time RT-PCR analysis of the Foxp3 mRNA expression in pancreas, splenocytes and PDLN of vehicle or MLDSTZ treated mice was performed as described in the Methods section. Results are expressed as means ± SEM, from two experiments (n = 3 mice/group/experiment). Unpaired t-tests (a, c, e; spleen and PDLNs) and Wilcoxson Rank Sum tests (e; pancreas) were performed for comparisons between vehicle and MLDSTZ treated groups on corresponding days. *, ** and *** denote p < 0.05, p < 0.01, and p < 0.001, respectively.

Figure 2. Impaired production of IL-10 and IL-35 by Treg cells in MLDSTZ induced T1D.

(A) Concentrations of IL-10 and (B) IL-35 in supernatants of stimulated CD4⁺CD25⁺ Treg cells sorted from thymic glands, PDLNS and spleen of vehicle or MLDSTZ treated mice. (C) Representative histograms showing the expression of IL-12p35 (left panels) or Ebi3 (right panels) in CD4⁺CD25⁺Foxp3⁺ Treg or CD4⁺CD25⁻ Th cells in thymocytes, PDLNs cells and splenocytes of vehicle or MLDSTZ injected mice on indicated days. Results are expressed as means ± SEM, from two experiments (n = 3 mice/group/experiment). Unpaired t-tests were performed for comparisons between vehicle and MLDSTZ treated groups on corresponding days. *, ** and *** denote p < 0.05, p < 0.01, and p < 0.001, respectively.

Figure 3. Treg cells acquire a T effector cell phenotype in MLDSTZ T1D.

(A) The percentage of IFN-γ⁺ cells in Foxp3⁺ Treg cells of thymic glands, PDLNs and spleen of vehicle and MLDSTZ injected mice (gating strategies are shown in Supplementary Fig. 5A). (B) Representative histograms showing the expression of Eos, IL-2, and IL-17 in CD4⁺CD25⁺Foxp3⁺ Treg cells. Results are expressed as means ± SEM, from two experiments (n = 3 mice/group/experiment). Unpaired t-tests were performed for comparisons between vehicle and MLDSTZ treated mice. *, ** and *** denote p < 0.05, p < 0.01, and p < 0.001, respectively.

Figure 4. Increased numbers of tTreg and pTreg cells can not keep the numbers of Th17 cells down in MLDSTZ induced T1D.

(A) CD4⁺CD25⁻ Th cells were analyzed by flow cytometry in thymic glands, PDLNs and spleen of vehicle and MLDSTZ injected mice. The proportions of CD4⁺CD25⁻ Th cells in total was expressed as means ± SEM. All significant differences in PDLN cells are not indicated in the figure, but they are, however, described in the Results section. (B) The percentage of IL-17a⁺ T cells in CD4⁺CD25⁻ Th cells were analyzed by flow cytometry in thymic glands, PDLNs and spleen of vehicle and MLDSTZ injected mice. (C) Representative histograms showing the expression of IL-17a in CD4⁺CD25⁻ Th cells. (D) Relative IL-17a mRNA expression in PDLN cells (on day 10) and splenocytes (on day 21) of vehicle or MLDSTZ injected mice were analyzed by real time RT-PCR. Results are expressed as means ± SEM, from two experiments (n = 3 mice/group/experiment). Unpaired t-tests were performed for comparisons between vehicle and MLDSTZ treated groups on corresponding days. ** and *** denote p < 0.01, and p < 0.001, respectively.

Figure 5. IL-35 prevents induction of MLDSTZ induced T1D and reverses established MLDSTZ induced T1D.

(A) MLDSTZ treated mice were injected with mouse recombinant IL-35 (0.75 μg/day, i.p.) or 200 μl PBS/day i.p. from day 6 for 8 days (shaded area). Six mice from the control group and six mice from the IL-35 treated group were sacrificed on day 14. After discontinuing IL-35 treatment, one mouse became hyperglycemic on day 26, while five mice remained normoglycemic. (B) MLDSTZ treated mice that had been hyperglycemic for two consecutive days (on day 10, after the first injection of STZ) were treated with mouse recombinant IL-35 for 8 days (shaded area) (n = 4 mice). Representative images of (C) insulitis grade 3 (>2/3 of the islet area infiltrated with mononuclear cells) and insulin grade 1 (<25% of insulin-positive cells). (D) Insulitis grade 2 (1/3-2/3 of the islet area infiltrated with mononuclear cells) and insulin grade 2 (25–50% insulin-positive cells). (E) Insulitis grade 0 (no infiltration in islets) and insulin grade 3 (>75% insulin-positive cells). (F) Haematoxylin and eosin stained sections of pancreata of IL-35 (0.75 μg daily, i.p.) or PBS (200 μl, i.p.) treated mice were analyzed for insulitis by scoring (0, 1, 2, 3 and 4) as described in the Methods section. (G) Insulin stained sections of pancreata of MLDSTZ + PBS or MLDSTZ + IL-35 treated mice were analyzed for insulin by scoring (0, 1, 2, and 3) as described in the Methods section. ^#MLDSTZ treated mice received IL-35 (0.75 μg daily, i.p.) for 8 days (starting from day 6 after the first injection of STZ). From day 14 (after the first injection of STZ) the IL-35 treatment was discontinued and mice were sacrificed on day 30 to analyze the morphology of pancreata. ^§MLDSTZ treated mice that had been hyperglycemic for two consecutive days (on day 10) were treated with mouse recombinant IL-35 (0.75 μg daily, i.p.) for 8 days as indicated in Fig. 4B. The mice were sacrificed on day 32. Mann-Whitney Rank Sum tests were performed for comparisons between MLDSTZ + PBS and MLDSTZ + IL-35 treated groups.

Figure 6. IL-35 administration increased the numbers of iT_R35 cells and production of IL-10 in prevention of development of MLDSTZ induced T1D.

Flow cytometry histograms showing the expression of Ebi3 in (A) CD4⁺CD25⁻ T cells or (B) CD4⁺CD25⁺Foxp3⁺ Treg cells of thymus, PDLNs cells and spleen of MLDSTZ + PBS or MLDSTZ + IL-35 treated mice on day 14. (C) The IL-10 concentration was determined in the serum of different groups of MLDSTZ + PBS and MLDSTZ + IL-35 treated mice by using ELSA assay. (D) Histograms showing the expression of CD39 in CD4⁺CD25⁺Foxp3⁺ Treg cells of thymus, PDLNs cells and spleen of MLDSTZ + PBS or MLDSTZ + IL-35 treated mice on day 14. Results are expressed as means ± SEM (n = 6), from two experiments (n = 3 mice/group/experiment). One-way ANOVA followed by Shapiro-Wilk tests (C) were performed for comparisons between IL-35 and PBS treated mice, ** denotes p < 0.01.

Figure 7. IL-35 administration reverses the phenotypic shift of the Treg cells in MLDSTZ induced T1D.

The proportions of Tbet⁺ cells in (A) CD4⁺CD25⁺Foxp3⁺, (B) CD4⁺CD25⁺Foxp3⁺Helios⁺, and in (C) CD4⁺CD25⁺Foxp3⁺Helios⁻ Treg cells were analyzed in thymus, PDLNs, and spleen of MLDSTZ + PBS or MLDSTZ + IL-35 treated mice on day 14 by flow cytometry. (D–F) The proportions of IL-17a⁺ cells in (D) CD4⁺CD25⁺Foxp3⁺, (E) CD4⁺CD25⁺Foxp3⁺Helios⁺ and in (F) CD4⁺CD25⁺Foxp3⁺Helios⁻ Treg cells were analyzed by flow cytometry in thymus, PDLNs, and spleen of MLDSTZ + PBS or MLDSTZ + IL-35 treated mice on day 14. (G) Representative histograms are showing the expression of Eos (left panel), mean fluorescent intensity (MFI) of Eos in CD4⁺CD25⁺Foxp3⁺ Treg cells (middle panel), and Bcl-2 (right panel) in CD4⁺CD25⁺Foxp3⁺ Treg cells of PDLNs of MLDSTZ + IL-35 and MLDSTZ + PBS treated mice. (H) CD4⁺CD25⁺Foxp3⁺Eos⁻ Treg cells were analyzed in PDLNs of MLDSTZ + PBS or MLDSTZ + IL-35 treated mice on day 14 by flow cytometry. Results are expressed as means ± SEM, from two experiments (n = 3 mice/group/experiment). Unpaired t-tests were performed for comparisons between IL-35 and PBS treated mice and *, ** and *** denote p < 0.05, p < 0.01and p < 0.001, respectively.

Figure 8. IL-35 administration reverses already established T1D in the NOD mouse model.

(A) NOD mice that had been hyperglycemic (>11.1 mM) for two consecutive days (new-onset diabetic) were treated with PBS (red) or mouse recombinant IL-35 (0.75 μg daily, i.p.) (black) for 8 days as indicated. Blood glucose levels were followed for one month. (B) The percentage of diabetic mice remaining normoglycemic after the treatment with PBS or IL-35, ** denotes p = 0.002. Fisher’s exact test was performed for comparisons on day 5. (C) Representative images of the islets of PBS or IL-35 treated NOD mice. Results are expressed as means ± SEM (n = 6).

Figure 9. IL-35 treatment maintains the phenotype of Treg cells in vitro.

(A) Thymocyte, PDLN cells and splenocytes were first gated for CD4, CD25 and Foxp3 expression, and then CD4⁺CD25⁺Foxp3⁺ Treg cells were further gated for IL-17a expression. Dot plots show representative experiments of PBS and IL-35 treated PDLNs cells of diabetic NOD mice. (B) The percentage of IL-17a⁺ cells in Foxp3⁺ Treg cells of thymic glands, PDLNs and spleen of PBS and IL-35 treated. (C) Thymocyte, PDLN cells and splenocytes were first gated for CD4, CD25 and Foxp3 expression, and then CD4⁺CD25⁺Foxp3⁺ Treg cells were further gated for Eos expression. Histogram shows representative experiments of PBS and IL-35 treated PDLNs cells of NOD diabetic mice. (D) The MFI of Eos in Foxp3⁺ Treg cells of thymic glands, PDLNs and spleen of PBS and IL-35 treated. Results are expressed as means ± SEM, from two experiments (n = 3 mice/group/experiment). Unpaired t-tests for comparisons between IL-35 and PBS treated groups were performed and *, ** and *** denote p < 0.05, p < 0.01and p < 0.001, respectively.

Figure 10. The circulating level of IL-35 is decreased in both new onset and long standing human T1D.

Circulating IL-35 levels in healthy controls (n = 13), recent onset T1D (n = 8) and long-standing human T1D patients (n = 19). One-way ANOVA followed by Tukey’s test was performed for comparisons and **, denote p < 0.01.

Figure 11. Tentative outline of IL-35 mediated protection against T1D in MLDSTZ.

In MLDSTZ treated mice Treg cells are increased in numbers, but these Treg cells do not protect against hyperglycemia. This might be due to a phenotypic shift of the Treg cells leading to expression of IFN-γ, IL-2, and IL-17. The acquisition of a Th1 and Th17 like phenotype by the Treg cells may be caused by an impaired expression of Eos. This further leads to a decreased production of anti-inflammatory cytokines (IL-10, IL-35, and TGF-β) and a failure to keep the numbers of cytotoxic CD8⁺ (Tc), Th1 and Th17 cells down and ultimately β-cell destruction. IL-35 administration reversed the phenotypic shift of the Treg cells, possibly by inducing the expression of Eos and maintained the regulatory phenotype of Treg cells to both prevented from development, and reversed established T1D.