The E3 ubiquitin ligase Itch regulates expression of transcription factor Foxp3 and airway inflammation by enhancing the function of transcription factor TIEG1

Abstract

Transforming growth factor-beta (TGF-beta) signaling in naive T cells induces expression of the transcription factor Foxp3, a 'master' regulator of regulatory T cells (T(reg) cells). However, the molecular mechanisms leading to Foxp3 induction remain unclear. Here we show that Itch-/- T cells were resistant to TGF-beta treatment and had less Foxp3 expression. The E3 ubiquitin ligase Itch associated with and promoted conjugation of ubiquitin to the transcription factor TIEG1. Itch cooperated with TIEG1 to induce Foxp3 expression, which was reversed by TIEG1 deficiency. Functionally, 'TGF-beta-converted' T(reg) cells generated from TIEG1-deficient mice were unable to suppress airway inflammation in vivo. These results suggest TIEG and Itch contribute to a ubiquitin-dependent nonproteolytic pathway that regulates inducible Foxp3 expression and the control of allergic responses.

Figure 1

Itch^−/− CD4⁺CD25⁻ cells are resistant to TGF-β-mediated suppression. (a) Flow cytometry of CFSE dilution in Itch^+/+ and Itch^−/− CD4⁺CD25⁻ responder cells labeled with CFSE, cultured with Itch^+/+ or Itch^−/− CD4⁺CD25⁺ T_reg cells and irradiated T cell–depleted splenocyte samples, then stimulated with anti-CD3 (1 μg/ml) and cultured for 72 h. MFI, mean fluorescence intensity. (b) The inhibitory function of T_reg cells, analyzed by measurement of ³[H]thymidine incorporation by CD4⁺CD25⁻ cells and CD4⁺CD25⁺ cells cultured together as described in a. (c) Proliferation of CD4⁺CD25⁻ cells isolated from Itch^+/+ and Itch^−/− mice, stimulated with anti-CD3 plus anti-CD28 and various concentrations of recombinant TGF-β and cultured for 72 h, analyzed by measurement of ³[H] thymidine incorporation. Data are representative of three independent experiments (mean ± s.d., b,c).

Figure 2

Itch regulates Foxp3 expression in TGF-β-treated CD4⁺CD25⁻ cells. (a) Real-time PCR analysis of Foxp3 mRNA in CD4⁺CD25⁻ cells isolated from Itch^+/+ and Itch^−/− mice and stimulated with anti-CD3 plus anti-CD28 with (+) or without (−) TGF-β (5 ng/ml). Arbitrary units (AU) are normalized to the abundance of Hprt1 mRNA. (b) Immunoblot analysis of Foxp3 expression in an experiment similar to that in a; Foxp3 expression is normalized to that of β-actin. (c) CSFE dilution analysis of the inhibitory function of TGF-β-treated Itch^+/+ or Itch^−/− CD4⁺ T cells cultured together with CD4⁺CD25⁻ responder cells isolated from Itch^+/+ mice. None, Itch^+/+ responder cells cultured alone. (d) Flow cytometry of Itch^+/+ and Itch^−/− CD4⁺CD25⁻ cells stimulated for 6 d with anti-CD3 plus anti-CD28 and TGF-β; live cells were then restimulated for 12 h with anti-CD3 plus anti-CD28, with brefeldin A added during the final 2 h, then cells were stained intracellularly with fluorescein isothiocyanate–labeled anti-IL-4 (FITC–anti-IL-4) and phycoerythrin-labeled anti-Foxp3 (PE–anti-Foxp3). Numbers in quadrants indicate percent Foxp3⁺IL-4⁻ cells (top left) or Foxp3⁻ IL-4⁺ cells (bottom right). (e) Immunoblot of lysates of Itch^+/+ and Itch^−/− CD4⁺CD25⁻ cells stimulated with anti-CD3 plus anti-CD28 and TGF-β with or without anti-IL-4 (10 μg/ml), analyzed with anti-Foxp3, then reprobed with anti-Grb2 (loading control). Data are representative of three independent experiments.

Figure 3

Itch associates with TIEG1 and targets it for ubiquitination. (a) Immunoblot analysis of Smad2 phosphorylation in lysates of CD4⁺CD25⁻ cells sorted from Itch^+/+ and Itch^−/− mice and stimulated with anti-CD3 plus anti-CD28 with or without TGF-β (5 ng/ml), analyzed with antibody to phosphorylated Smad2 (pSmad2). Total Smad2-Smad3 (Smad2/3) and Grb2 serve as loading controls. (b) Immunoblot of lysates of CD4⁺CD25⁻ cells precipitated with GST alone or GST-TIEG1; precipitates (top) and total cell lysates (CL; bottom) were analyzed with anti-Itch. (c) Immunoblot analysis of CD4⁺CD25⁻ cells left untreated or treated with TGF-β; lysates were precipitated with GST alone or GST-TIEG1, and precipitates (top) and total cell lysates (bottom) were analyzed with anti-Itch. (d) Immunoblot analysis of CD4⁺CD25⁻ cells treated with TGF-β; lysates were precipitated with GST alone or GST-WW and analyzed with anti-TIEG1. (e) Immunoassay of the in vivo association between Itch and TIEG1 in TGF-β-stimulated T cells; lysates were immunoprecipitated with normal goat serum (NGS), anti-TIEG1 (α-TIEG1), normal mouse serum (NMS) or anti-Itch (α-Itch) and were analyzed by immunoblot with anti-TIEG1 (top) or anti-Itch (bottom). (f) Immunoassay of 293T cells transiently transfected with various plasmids (above lanes), then lysed 48 h later and denatured in 1% SDS; lysates were immunoprecipitated (IP) with anti-Myc and analyzed by immunoblot (IB) with anti-hemagglutinin (α-HA). Left margin, molecular sizes in kilodaltons (kDa); right margin, positions of monoubiquitin (Mono-Ub) and polyubiquitin (Poly-Ub). Middle, same membrane reprobed with anti-Myc. Bottom, immunoblot analysis of aliquots of cell lysates with anti-Itch. HA-Ub, hemagglutinin-tagged ubiquitin. (g) Immunoassay of 293T cells transfected with various plasmids (above lanes) plus plasmid encoding an active form of TGF-β receptor 1 (TβR1*). Right margin, positions of monoubiquitin, diubiquitin (+2), triubiquitin (+3), quadriubiquitin (+4) and polyubiquitin. (h) Immunoprecipitation and immunoblot analysis of transfected 293T cells separated into cytoplasmic (C) and nuclear (N) fractions; GAPDH (glyceraldehyde phosphate dehydrogenase) and lamin serve as ‘markers’ for the cytoplasmic and nuclear fractions, respectively. Data are representative of three independent experiments.

Figure 4

Itch-mediated ubiquitination of TIEG1 is necessary for Foxp3 expression. (a) Foxp3 promoter activity in Jurkat T cells transfected with the pGL3-Foxp3-Ce reporter plasmid, control vector, Itch, Itch-CA and/or TIEG1, then treated for 24 h with TGF-β, analyzed by measurement of luciferase (luc) activity (normalized to that of β-galactosidase). (b) EMSA of DNA-protein binding in nuclear extracts of untreated or TGF-β-treated CD4⁺CD25⁻ cells mixed with radiolabeled oligonucleotide probe corresponding to a GC-rich region in the Foxp3 promoter; anti-TIEG1 or normal goat serum was used for higher-order antibody-protein complex formation ‘supershifting’. Right margin, antibody-dependent ‘supershifted’ bands. (c) EMSA of DNA binding in nuclear lysates from untreated or TGF-β-treated Itch^+/+ and Itch^−/− CD4⁺CD25⁻ cells. (d) Chromatin immunoprecipitation assay of Itch^+/+ and Itch^−/− CD4⁺CD25⁻ cells; chromatin DNA obtained before (Input) and after (IP) immunoprecipitation with anti-TIEG1 or control normal goat serum was analyzed by PCR with primers specific for the Foxp3 promoter. (e) ‘Supershift’ assay, as described in b, of 293T cells transfected with various plasmids (above lanes), analyzed with anti-hemagglutinin. mIgG, mouse IgG. (f) Immunoblot analysis of Itch^+/+ and Itch^−/− CD4⁺CD25⁻ cells transduced with retrovirus vector pMIG-Itch or pMIG-Itch-CA; GFP⁺ cells sorted by flow cytometry were stimulated with anti-CD3 plus anti-CD28 and TGF-β and analyzed with anti-Foxp3. Bottom, same membrane probed with anti-Grb2 (loading control). (g) Chromatin immunoprecipitation assay of the interaction of TIEG1 with the Foxp3 promoter in cells reconstituted as described in f. Data are representative of three repeated experiments.

Figure 5

TIEG1 expression restores Foxp3 expression in Itch^−/− T cells. (a) Immunoblot analysis of TIEG1, Smad2 and Smad4 in lysates of Itch^+/+ and Itch^−/− CD4⁺CD25⁻ cells stimulated with anti-CD3 plus anti-CD28 with or without TGF-β (5 ng/ml). (b) Flow cytometry of CD4⁺CD25⁻ cells transduced with pMX or pMX-TIEG1, then stained intracellularly with phycoerythrin-labeled anti-Foxp3 after 6 d. Numbers in quadrants indicate percent Foxp3⁺GFP⁻ cells (top left) or Foxp3⁺GFP⁺ cells (top right). (c) Immunoblot analysis of Foxp3 in lysates of CD4⁺CD25⁻ cells transduced with pMX or pMX-TIEG1; GFP⁺ cells were sorted by flow cytometry and treated with TGF-β. Bottom, same lysates probed with anti-Grb2 (loading control). Data are representative of three independent experiments.

Figure 6

TIEG1-deficient T cells fail to express Foxp3. (a) [³H]thymidine incorporation by CD4⁺CD25⁻ cells from TIEG1–wild-type mice (open bars) and TIEG1-deficient mice (filled bars), stimulated for 72 h with anti-CD3 plus anti-CD28 and various concentrations of TGF-β (horizontal axis) and pulsed for the final 12 h with [³H]thymidine. (b) Flow cytometry of Foxp3 expression by CD4⁺CD25⁻ cells from TIEG1–wild-type mice (TIEG1 WT) and TIEG1-deficient mice (TIEG1 MUT), stimulated for 7 d with anti-CD3 plus anti-CD28 with or without TGF-β (5 ng/ml), analyzed by intracellular staining. Numbers in the top right quadrants indicate percent Foxp3⁺CD4⁺ T cells. (c) CFSE dilution by CD4⁺CD25⁻ cells isolated from TIEG1–wild-type mice, labeled with CFSE and cultured together with TGF-β-treated TIEG1–wild-type or TIEG1-deficient T cells. Data are from one of three experiments (mean ± s.d. of triplicate samples, a).

Figure 7

Reconstitution of TIEG1 restores Foxp3 expression. (a) Flow cytometry of CD4⁺CD25⁻ cells sorted from TIEG1–wild-type or TIEG1-deficient mice, transduced with retroviral vector pMX (control) or pMX-TIEG1, then treated with TGF-β and stained intracellularly with anti-Foxp3. Numbers in quadrants indicate percent Foxp3⁺GFP⁻ cells (top left) or Foxp3⁺GFP⁺ cells (top right). (b) PKH26 dilution by wild-type (WT), TIEG1-deficient and Itch^−/− CD4⁺CD25⁻ cells transduced with pMX or pMX-TIEG1; GFP⁺ cells sorted by flow cytometry (responder T cells) were cultured together with CD4⁺CD25⁺ T_reg cells isolated from wild-type mice to assess the responsiveness of the responder cells to T_reg cell–mediated suppression. Data represent one of three independent experiments.

Figure 8

TIEG1-deficient T_reg cells are defective in inhibiting airway inflammation. CD4⁺CD25⁻ cells sorted from TIEG1–wild-type and TIEG1-deficient mice (n = 4 mice per group) were transduced with pMX or pMX-TIEG1 retrovirus and left untreated or treated with TGF-β; GFP⁺ cells (2.5 × 10⁶) were sorted by flow cytometry and then adoptively transferred into naive wild-type mice (n = 4 mice per group). Recipient mice were immunized, followed by intranasal challenge with aerosolized OVA. (a) Lung histology (hematoxylin and eosin staining). Original magnification, ×00. (b,c) IL-4 concentration (b) and eosinophils (c) in bronchoalveolar lavage fluid. (d) Enzyme-linked yimmunosorbent assay of serum IgE titers. Arbitrary units were converted to a known IgE standard. (b–d) Open bars, TIEG1–wild-type; filled bars, TIEG1-deficient. Data are representative of three independent experiments (mean ± s.d. of three samples, b–d).