IL-9 induces differentiation of TH17 cells and enhances function of FoxP3؉ natural regulatory T cells Wassim Elyamana, Elizabeth M. Bradshawa, Catherine Uyttenhoveb, Vale´ rie Dardalhona, Amit Awasthia, Jaime Imitolaa, Estelle Bettellia, Mohamed Oukkaa, Jacques van Snickb, Jean-Christophe Renauldb, Vijay K. Kuchrooa, and Samia J. Khourya,1 aCenter for Neurologic Diseases, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA 02115 and bLudwig Institute for Cancer Research and Experimental Medicine Unit, Universite´Catholique de Louvain, Brussels, Belgium Edited by Richard A. Flavell, Yale University School of Medicine, New Haven, CT, and approved March 31, 2009 (received for review December 11, 2008) The development of T helper (TH)17 and regulatory T (Treg) cells is ing large amounts of IL-10, but induce severe colitis and peripheral reciprocally regulated by cytokines. Transforming growth factor neuritis upon adoptive transfer into immune deficient hosts (14). ؉ ␤ (TGF)- alone induces FoxP3 Treg cells, but together with IL-6 or IL-21 Another recent report describes reprogramming of TH2 cells by induces TH17 cells. Here we demonstrate that IL-9 is a key molecule TGF-␤ to cells producing IL-9 and IL-10 (11), and these cells were that affects differentiation of TH17 cells and Treg function. IL-9 pre- also shown to have effector and not regulatory functions. These ␤ dominantly produced by TH17 cells, synergizes with TGF- 1 to differ- data suggest that IL-9/IL-10-producing T cells are not regulatory T ؉ entiate naïve CD4 T cells into TH17 cells, while IL-9 secretion by TH17 cells but effector T cells that induce tissue inflammation. However, cells is regulated by IL-23. Interestingly, IL-9 enhances the suppressive whether these effector functions are mediated by IL-9 and what role ؉ ؉ functions of FoxP3 CD4 Treg cells in vitro, and absence of IL-9 does IL-9 play in pathogenicity of this unique subset of T cells has signaling weakens the suppressive activity of nTregs in vivo, leading not been evaluated. to an increase in effector cells and worsening of experimental auto- In the present study, we show that TH17 cells produce large immune encephalomyelitis. The mechanism of IL-9 effects on TH17 quantities of IL-9 that act on both TH17 and Treg cells. In the and Tregs is through activation of STAT3 and STAT5 signaling. Our presence of TGF-␤, IL-9 differentiates naïve CD4ϩ T cells into findings highlight a role of IL-9 as a regulator of pathogenic versus TH17 cells. Paradoxically, IL-9 also acts on nTregs and enhances their protective mechanisms of immune responses. suppressive function in vitro and in vivo. Mice lacking IL-9 receptor (IL-9RϪ/Ϫ) exhibit a more severe course of experimental autoim- ͉ ͉ autoimmunity regulatory cells tolerance mune encephalomyelitis (EAE) and have a defect in the suppressive activity of Tregs. n the presence of antigen stimulation, naïve CD4ϩ T cells Iproliferate and differentiate into T helper type 1 (TH1) cells, TH2 Results cells, or interleukin-17 (IL-17)-producing T helper cells (TH17 cells) IL-9 Is Produced by TH17 Cells. TH17 cells produce several cytokines to exert specific effector functions. TH17 cells express the transcrip- including IL-17, IL-21, and IL-22 (2). We sought to determine the ␥ tion factor retinoic acid orphan receptor gamma (ROR- )t (1), additional cytokines that are secreted by TH17 cells. FACS-sorted participate in the control of extracellular pathogens, and have an naïve CD4ϩCD62LhiFoxP3Ϫ T cells isolated from FoxP3.GFP important role in human and experimental autoimmunity (2). TH17 ‘‘knock-in’’ (FoxP3.GFP.KI) reporter mice were differentiated into cells have been identified as major inducers of tissue inflammation TH1, TH2, and TH17 according to established protocols (Fig. S1A). and autoimmunity. Since exaggerated responses of TH1, TH2, and Inducible Tregs(iTregs) (7) were generated by activating TH17 cells can induce tissue inflammation, the maintenance of CD4ϩCD62LhiFoxP3Ϫ T cells in the presence of TGF-␤ (Fig. S1B), immune homeostasis and prevention of immunopathology is me- and nTregs were sorted from naïve FoxP3.GFP.KI mice based on in diated by subsets of T cells called regulatory T cells (Tregs). Treg cell vivo FoxP3 expression (Fig. S1C). Using multiplex bead-based differentiation and function are driven by the transcription factor Luminex technology, we identified IL-9 as a dominant cytokine forkhead box P3 (FoxP3) (3, 4), and they are closely related to the ␤ produced both by TH2 and TH17 cells (Fig. 1A). These data were generation of TH17 cells. Transforming growth factor (TGF)- 1 confirmed by quantitative Taqman PCR (Fig. 1A) and by flow induces the differentiation of T cells (5), whereas TGF-␤1in reg cytometry (Fig. 1B). Moreover, IL-9 was not detected in the combination with IL-6 (6, 7) or IL-21 (8) results in the differenti- supernatants of nT or iT (Fig. 1A), but IL-9R mRNA was ation of T 17 cells. regs reg H highly expressed by nT and not by iT (Fig. 1A). IL-9 is a T cell-derived factor preferentially expressed by T 2 regs regs H Naïve CD4ϩ T cells activated with plate-bound anti-CD3 and cells (9), although a previous study suggested that regulatory T cells anti-CD28 in the presence of recombinant mouse IL-9 had in- produce more IL-9 than TH2 cells (10); however this was not creased secretion of the TH2 (IL-4, IL-10, and IL-13), cytokines, confirmed in induced (iTregs) or natural (nTregs) (11). It is a ␥ pleiotropic cytokine that targets cells of the lymphoid, myeloid, and and at the same time down-regulated IFN (IFN)- and granulocyte- mast cell lineages, as well as lung epithelial cells. IL-9 activities are macrophage colony stimulation factor (GM-CSF) production (Fig. mediated by a specific IL-9 receptor chain that forms a het- IMMUNOLOGY erodimeric receptor with the common gamma chain (␥c) also Author contributions: W.E., C.U., J.I., J.V.S., J.-C.R., and S.J.K. designed research; W.E., involved in IL-2, 4, 7, 15, and 21 signaling. The IL-9 receptor and E.M.B., C.U., V.A.D., A.A., and J.V.S. performed research; C.U., E.B., M.O., J.V.S., J.-C.R., and ␥c associate with Janus kinase (JAK)1 and JAK3, respectively, and V.K.K. contributed new reagents/analytic tools; W.E., E.M.B., C.U., V.A.D., and J.V.S. ana- trigger the activation of STAT1, 3, and 5 (12, 13). Dysregulated IL-9 lyzed data; and W.E. wrote the paper. response in vitro can lead to autonomous cell growth and malignant The authors declare no conflict of interest. transformation of lymphoid cells associated with constitutive acti- This article is a PNAS Direct Submission. vation of the JAK/STAT pathway (12). We have recently described 1To whom correspondence should be addressed. E-mail: [email protected]. a subset of T cells that predominantly produces IL-9 together with This article contains supporting information online at www.pnas.org/cgi/content/full/ IL-10 and do not exhibit any regulatory properties despite produc- 0812530106/DCSupplemental. www.pnas.org͞cgi͞doi͞10.1073͞pnas.0812530106 PNAS ͉ August 4, 2009 ͉ vol. 106 ͉ no. 31 ͉ 12885–12890 Downloaded by guest on October 1, 2021 A B Th1 Th2 Th17 0.12 0.18 1.13 0.64 3.56 1.04 5 5 5 iTregs 10 10 10 nTregs 4 10 104 104 Th17 9 3 - 10 103 103 Th2 L I 2 Th1 10 102 102 Th0 0 0 0 67.6 32.1 84.2 13.9 82.8 12.6 0 50 100 150 200 250 300 350 400 450 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0 1 2 3 4 5 6 7 8 9 0 102 103 104 105 0102 103 104 105 0102 103 104 105 IL-9 (pg/ml) IL-9 relative expression IL-9R relative expression IFN-γ IL-10 IL-17 TGF-β1 TGF-β1 C IL-4 IFN-γ 200 TGF-β 500 *** IL-21 400 IL-9 IL-6 300 100 IL-6 + TGF-β 200 IL-21 + TGF-β IL9 + TGF-β 100 No cytokine 0 0 0 100 200 300 0 100 200 300 400 500 no cytokine IL-9 no cytokin IL-9 pg/ml pg/ml ϩ hi Ϫ Fig. 1. TH17 cells produce IL-9. CD4 CD62L FoxP3/GFP T cells from FoxP3-GFP.KI mice were stimulated with anti-CD3 and anti-CD28 for 3–5 days in the presence of the corresponding cytokines. (A) IL-9 relative expression as determined by bead-based Luminex assay (left) and by quantitative RT-PCR (middle) and IL-9 ϩ hi receptor relative expression (right) in TH0, TH1, TH2, TH17, nTregs, and iTregs.(B) Intracellular staining of cytokines in naïve CD4 CD62L T cells from B6 mice sorted ϩ by flow cytometry and cultured for 4 days in TH1, TH2, or TH17 conditions. (C) Effects of IL-9 on the production of IL-4, IFN-␥, and TGF-␤1 by naïve CD4 T cells as measured by bead-based Luminex assay. For TGF-␤1 assay, supernatants were ‘‘acid-treated’’ followed by bead-based Luminex assay. Effect of IL-9 on TGF-␤1 expression in T cells was also measured by quantitative Taqman PCR. 1C and Fig. S2). Interestingly, cells exposed to IL-9 produce IL-9 might normally contribute to TH17 differentiation even when significant amounts of TGF-␤1 and this was confirmed by Taqman TH17 cells were differentiated by other cytokine cocktails (TGF-␤ PCR (Fig.