Antigen receptor-mediated depletion of FOXP3 in induced regulatory T-lymphocytes via PTPN2 and FOXO1

Abstract

Regulatory T-cells induced via IL-2 and TGFβ in vitro (iTreg) suppress immune cells and are potential therapeutics during autoimmunity. However, several reports described their re-differentiation into pathogenic cells in vivo and loss of their key functional transcription factor (TF) FOXP3 after T-cell antigen receptor (TCR)-signalling in vitro. Here, we show that TCR-activation antagonizes two necessary TFs for foxp3 gene transcription, which are themselves regulated by phosphorylation. Although the tyrosine phosphatase PTPN2 is induced to restrain IL-2-mediated phosphorylation of the TF STAT5, expression of the TF FOXO1 is downregulated and miR-182, a suppressor of FOXO1 expression, is upregulated. TGFβ counteracts the FOXP3-depleting TCR-signal by reassuring FOXO1 expression and by re-licensing STAT5 phosphorylation. Overexpressed phosphorylation-independent active versions of FOXO1 and STAT5 or knockdown of PTPN2 restores FOXP3 expression despite TCR-signal and absence of TGFβ. This study suggests novel targets for stabilisation and less dangerous application of iTreg during devastating inflammation.

Figure 1. TCR-signalling suppresses FOXP3 expression in iTregs but not ex vivo purified Tregs.

(a–f): intracellular staining of FOXP3; numbers indicate percentages of FOXP3⁺ cells according to the indicated range gate settings. (a) iTregs generated by 72 h stimulation via CD3/28 plus IL-2/TGFβ or GFP⁺ CD4⁺ nTegs sorted from DEREG mice, respectively. (b–e) re-stimulation of the cells depicted in (a) for 48 h by anti-CD3 or PMA/Ionomycin (PI), in the presence of IL-2. (c,e) Statistical evaluation plus s.d. (Student's t-test) of five (b,c) or three (d,e) consecutive experiments. ***P<0.001. (f) Ly 5.1+ iTregs re-cultured with IL-2 and Ly 5.1 negative congenic APC, with or without SEB, stained with anti-Ly5.1, anti-Vβ8 or anti-Vβ6 and gated for Ly5.1⁺ cells. Two experiments with similar outcome. ns: not significant.

Figure 2. TCR-signalling interferes with active FOXP3 production.

(a,b (top panels), c,d): intracellular staining of FOXP3 after 48 h of re-culture. (a) WT iTregs generated as before were re-cultured with IL-2, but without αCD3 and with or without CHX (10 μg ml⁻¹). (b): iTregs were generated from sorted GFP-negative CD4⁺ DEREG cells. After 72 h, GFP⁺ cells were again sorted to a purity of >95% (range gate settings as depicted in b), and re-cultured for 48 h with IL-2 and with or without αCD3 (b, lower panels: GFP expression). (a,b) Statistical evaluation plus s.d. (Student's t-test) of three consecutive experiments. (c,d) WT iTregs were re-cultured for 48 h with IL-2, αIL-2, TGFβ or IL-6 with or without αCD3, as indicated. (d) Baseline represents cells re-cultured with IL-2 only. Five (c) or three (d) experiments with similar outcome. Statistical evaluation ± s.d. of (c) in (e) and of (d) in (f). n.s.: not significant. **P< 0.01; ***P< 0.001.

Figure 3. TCR stimulation hampers IL-2 signalling.

(a) Western blot for pSTAT5, STAT5 and β-actin in lysates of iTregs re-cultured for 24 h under the indicated conditions, deprived of IL-2 for 1 h and resupplied with IL-2 (200 U ml⁻¹) for 10 min. (b,c) Staining of surface CD25, CD122 or CD132 (open areas) vs isotype controls (grey areas) or of FOXP3 on iTreg re-cultured for 48 h in the presence of IL-2 (b: 20 U ml⁻¹; c: 1 U ml⁻¹) with or without αCD3. (a) Four, (b,c) two experiments with similar outcome.

Figure 4. The p-Tyr phosphatase PTPN2 interferes with STAT5 phosphorylation.

(a) Experiments similar as described in Fig. 3a were performed, however IL-2 was deprived for 90 min with application of the indicated concentrations of Na₃VO₄ for the last 45 min and resupply of IL-2 (200 U ml⁻¹) for 10 min. Three experiments with similar outcome. (b) iTregs were induced as before, nucleofected with different siRNAs (one type per lane) directed against the indicated panel of p-Tyr phosphatases or with scrambled control siRNA and further incubated with IL-2 and without TCR-signal for 24 h, before being deprived of and resupplied with IL-2 as in A. Western blots were stained with antibodies to pSTAT5, total STAT5, PTPN2 or PTPN6. Numbers refer to relative values of densitometry. (c) western blot for PTPN2 in iTregs either immediately after 72 h induction (‘3D iTreg') or after 24 or 48 h re-culture with IL-2 and with or without αCD3. One representative of Figs 4b and 5c independent experiments.

Figure 5. TCR stimulation upregulates miR-182 and suppresses FOXO1.

(a) Western Blot for FOXO1 and β-actin in lysates of iTreg re-cultured for 24 h in the presence of IL-2 under the indicated conditions. Three experiments with similar outcome. (b) qRT–PCR analysis of miR-182 expression in iTreg re-cultured for 48 h with IL-2 under the indicated conditions. Statistical evaluation plus s.d. (Student's t-test) of four separate experiments. *P<0.05. ns: not significant.

Figure 6. Overexpression of CA STAT5 and CA FOXO1 and knockdown of PTPN2 overcome the negative TCR-signal.

(a) Primed iTregs were retrovirally transduced twice within 48 h with CA STAT5 or CA FOXO1 or the respective controls pMIG or pMIT, followed by 48 h of re-culture under the indicated conditions. Intracellular staining of FOXP3 in cells gated for successful double infection. Five experiments with similar outcome. Histograms from one representative experiment and summary (± s.d.) of the experiments with 5 μg ml⁻¹ of αCD3. (b,c). Primed iTregs were nucleofected with siRNA against PTPN2 followed by further 24 h of culture without the TCR-signal and 24 h of re-culture under the indicated conditions (in (c): PMA:5 ng ml⁻¹; αCD3: 0.5 μg ml⁻¹). Intracellular staining of FOXP3. (c) RFP⁺ cells were sorted after the induction period (purity 98-99%). (b,c) The histograms are representative for three experiemts summarized in the respective diagrams (± s.d.) (Student's t-test) depicted to the right. Baseline represents cells re-cultured with IL-2 only. *P<0.05; **P<0.01; ***P<0.001.

Figure 7. Antigen-specific downregulation of FOXP3 in iTregs in vivo.

iTregs were induced from OT II/FIR cells, were transfected with either CA FOXO1 or CA STAT5 or both, sorted for RFP⁺ cells (purity 96%) and were transferred i.p into C57BL/6 recipient mice with or without OVA and/or LPS (four mice per group). Mesenteric LN cells were analysed 3 days later by flow cytometry. (a): Gating strategy. (b) FOXP3 staining of cells within the lower left (‘double negative') gate defined in the third panel in (a). Columns depict the mean ± s.d. of FOXP3⁺ (%) double negative cells of the different groups of mice. (c) Mean ± s.d. (Student‘s t-test) of FOXP3⁺ cells (%) transfected with CA STAT5 or CA FOXO1 or both of the different groups of mice. Two experiments with similar outcomes; the P-values analogous to those shown here were <0.01 in the second experiment.

Figure 8. Cartoon summarizing the main findings.

.