Nonredundant roles for Stat5a/b in directly regulating Foxp3

Abstract

Stats (signal transducers and activators of transcription) regulate multiple aspects of T-cell fate. T regulatory (Treg) cells are a critical subset that limits immune responses, but the relative importance of Stat5a/b versus Stat3 for Treg cell development has been contentious. We observed that peripheral CD25(+)CD4(+) T cells were reduced in Stat5(DeltaN) mice; however, the levels of Foxp3, a transcription factor that is critical for Treg cells, were normal in splenic CD4(+) T cells even though they were reduced in the thymus. In contrast, complete deletion of Stat5a/b (Stat5(-/-)) resulted in dramatic reduction in CD25- or Foxp3-expressing CD4(+) T cells. An intrinsic requirement was demonstrated by reduction of Stat5a/b in CD4-expressing cells and by stem cell transplantation using Stat5(-/-) fetal liver cells. Stat5a/b were also required for optimal induction of Foxp3 in vitro and bound directly to the Foxp3 gene. Reduction of Stat3 in T cells did not reduce the numbers of Treg cells in the thymus or spleen; however, Stat3 was required for IL-6-dependent down-regulation of Foxp3. Therefore, we conclude that Stat5a/b have an essential, nonredundant role in regulating Treg cells, and that Stat3 and Stat5a/b appear to have opposing roles in the regulation of Foxp3.

Figure 1

Stat5a/b are critical for the generation of thymic Foxp3⁺ CD4⁺ T cells and maintenance of peripheral Foxp3⁺ CD4 T⁺ cells. CD25 and Foxp3 expression were analyzed by flow cytometry in CD4 SP thymocytes (A-C) and splenocytes (D-F) from 4- to 6-week-old Stat5^+/+ and Stat5^ΔN mice (A,D), Stat5^−/− mice (B,E), and Jak3^−/− and Il2rg^−/− mice (C,F). Numbers indicate the percentage of CD25⁺ or FoxP3⁺ cells delineated by the rectangles.

Figure 2

Intrinsic requirement of Stat5a/b for generation of Foxp3⁺ CD4 T cells. Irradiated C57BL/6 Rag2^−/− CD45.1⁺ congenic mice were reconstituted with 2 × 10⁶ total fetal liver cells from CD45.2⁺ WT (left panels) or Stat5^−/− (right panels) mice. After reconstitution (8 weeks), cell populations in the thymus (A) and spleen (B) were analyzed. Donor-derived CD4 SP T cells were analyzed for CD25 and Foxp3 expression. Numbers indicate the percentage of CD25⁺ or FoxP3⁺ cells delineated by the rectangles.

Figure 3

Reduction of thymic Foxp3⁺ CD4⁺ T cells with tissue-specific diminution of Stat5a/b levels. (A) CD25 and Foxp3 expression were assessed on sorted CD4 SP thymocytes from Stat5^fl/fl mice and YFP⁺CD4 SP thymocytes from Stat5^fl/−, CD4Cre, Yfp mice. (B) Average proportion of CD4 SP thymocytes. (C) Mean percentage of Foxp3⁺ CD4⁺ T cells. (D) Average total numbers of thymocytes. (E) Absolute numbers of Foxp3⁺ CD4⁺ T cells. Means ± SE (n = 6) are shown; *P < .01 as determined by Student t test. Numbers indicate the percentage of CD25⁺ or FoxP3⁺ cells delineated by the rectangles.

Figure 4

Reduction of peripheral Foxp3⁺ CD4⁺ T cells with tissue-specific decrease of Stat5a/b levels. (A) CD25 and Foxp3 expression were analyzed in sorted CD4⁺ SP splenocytes from Stat5^fl/fl mice and YFP⁺CD4⁺ SP splenocytes from Stat5^fl/−, CD4Cre, Yfp mice. Numbers indicate the percentage of CD25⁺ or FoxP3⁺ cells delineated by the rectangles. (B) Average proportion of CD4⁺ SP splenocytes. (C) Mean percentage of Foxp3⁺ CD4⁺ T cells. (D) Average total numbers of splenocytes. (E) Absolute numbers of Foxp3⁺ CD4 T cells. Means ± SE (n = 6) are shown (*P < .01).

Figure 5

Reduction of Stat3 levels in CD4 T cells has no effect on Foxp3 expression from both thymi and spleens. (A) Foxp3 expression was analyzed in CD4 SP thymocytes by flow cytometry. (B) The mean proportion of CD4 SP thymocytes, the mean percentage of Foxp3⁺ CD4⁺ T cells, the average total numbers of thymocytes, and the absolute numbers of Foxp3⁺ CD4 T cells in thymocytes are shown. (C) Foxp3 expression was analyzed in CD4⁺ splenocytes by flow cytometry. (D) The average proportion of CD4⁺ SP splenocytes, the mean percentage of Foxp3⁺ CD4⁺ T cells, the average total numbers of splenocytes, and the absolute numbers of Foxp3⁺ CD4 T cells are shown. Means ± SE (n = 5) are shown. In panels A and C, numbers indicate the percentage of FoxP3⁺ cells delineated by the rectangles.

Figure 6

Stat5a/b are important for the in vitro induction of Foxp3. (A) Sorted splenic CD25⁻CD4⁺ T cells from Stat5^fl/fl and YFP⁺ CD25⁻CD4⁺ T cells from Stat5^fl/−, CD4Cre, Yfp⁺ mice were cultured with plate-bound anti-CD3 and anti-CD28 combined with or without TGF-β1, IL-2 (100 U/mL), and/or anti–IL-2 antibody as indicated. Representative of 3 experiments. Numbers in quadrants indicate FoxP3⁺ or FoxP3⁻ cells. (B) Sorted splenic CD25⁻CD4⁺ T cells from Stat3^fl/fl and Stat3^fl/fl, MMTVcre mice were cultured with plate-bound anti-CD3 and anti-CD28 combined with or without TGF-β1 and IL-6 as indicated. Numbers indicate the percentage of FoxP3⁺ cells delineated by the rectangles.

Figure 7

Stat5 binds the Foxp3 gene. (A) Schematic of the mouse Foxp3 gene. Vertical lines depict potential Stat binding sites in the first intron and the putative promoter (I, II, III). Site IV does not contain a Stat-binding site and was used as a control. (B) Stat-binding sites in the mouse Foxp3 gene are underlined and aligned with sequences from other species. Site I is located between 006010267 and 006010275 in the mouse genome (http://genome.ucsc.edu/cgi-bin/hgGateway?hgsid = 83 436 504&clade = vertebrate&org = Mouse&Db = mm7). Site II is located between 006012112 and 006012120. Intronic sites designated III are located between 006017209 and 006017217, 006017406 and 006017414, and 006017523 and 006017531. All sequences are from the sense strand; note that the previous Stat-binding site sequences identified by Zorn et al were from the antisense strand, although the same sites were interrogated in our analysis. (C) Sorted thymic CD25⁺CD4 SP (■) and CD25⁻CD4 SP (▩) T cells were treated with IL-2 for 1 hour. Proteins and DNA were cross-linked with formaldehyde, cells were lysed, and DNA was sheared. Chromatin immunoprecipitation was performed using either normal rabbit serum or anti-Stat5 antibody. Quantification of immunoprecipitated DNA fragments was performed by real-time PCR using primers and probes for sites I, II, III, and the irrelevant site IV. Values were normalized to corresponding input control and are expressed as fold enrichment relative to normal rabbit serum for each experiment. Means ± SE are shown.