Interleukin-1 and IL-23 Induce Innate IL-17 Production from Gd T Cells, Amplifying Th17 Responses and Autoimmunity

Home , IL17A, Interleukin 17, Interleukin 23

Immunity Article

Interleukin-1 and IL-23 Induce Innate IL-17 Production from gd T Cells, Amplifying Th17 Responses and Autoimmunity

Caroline E. Sutton,1,3 Stephen J. Lalor,1,3 Cheryl M. Sweeney,1 Corinna F. Brereton,1 Ed C. Lavelle,2 and Kingston H.G. Mills1,* 1Immune Regulation Research Group 2Adjuvant Research Group School of Biochemistry and Immunology, Trinity College, Dublin 2, Ireland 3These authors contributed equally to this work *Correspondence: [email protected] DOI 10.1016/j.immuni.2009.08.001

SUMMARY IL-23p19-deficient (Il23aÀ/À) mice are resistant to the induction of experimental autoimmune encephalomyelitis (EAE) (Cua Th17 cells, CD4+ T cells that secrete interleukin-17 et al., 2003), and IL-23 was initially thought to be the critical cyto- (IL-17), are pathogenic in autoimmune diseases and kine in the differentiation of Th17 cells. However, naive murine their development and expansion is driven by the T cells do not express IL-23R and do not differentiate into cytokines IL-6, TGF-b, IL-21, IL-1, and IL-23. Th17 cells in response to IL-23 in vitro (Langrish et al., 2005; van Beelen et al., 2007). In contrast, IL-23 can expand Th17 cells However, there are also innate sources of IL-17. + Here, we show that gd T cells express IL-23R and from memory CD4 T cells (Aggarwal et al., 2003). We have also demonstrated that IL-1a or IL-1b synergizes with IL-23 to the transcription factor RORgt and produce IL-17, promote IL-17 production by murine T cells in the presence or IL-21, and IL-22 in response to IL-1b and IL-23, absence of T cell receptor engagement (Sutton et al., 2006). without T cell receptor engagement. IL-17-producing Furthermore, IL-17 production and development of EAE is gd T cells were found at high frequency in the brain of compromised in IL-1 receptor type I-defective (Il1r1À/À) mice. mice with experimental autoimmune encephalomy- IL-1 and IL-23 promote IL-17 production by memory but not elitis (EAE). gd T cells activated by IL-1b and IL-23 naive T cells (van Beelen et al., 2007). However, it has recently promoted IL-17 production by CD4+ T cells and been reported that IL-1 signaling is critical for Th17 cell differen- increased susceptibility to EAE, suggesting that gd tiation (Chung et al., 2009). Moreover, our studies, together with T cells act in an amplification loop for IL-17 produc- recent reports on human Th17 cells (Manel et al., 2008; Volpe tion by Th17 cells. Our findings demonstrate that gd et al., 2008) and on IL-23-driven innate IL-17 production (Hue T cells activated by IL-1b and IL-23 are an important et al., 2006; Takatori et al., 2009), suggest that the function of IL-23 may not be confined to expansion of memory Th17 cells. source of innate IL-17 and IL-21 and provide an alter- Although CD4+ T cells are considered to be the major source of native mechanism whereby IL-1 and IL-23 may IL-17, especially in autoimmune diseases, recent evidence has mediate autoimmune inflammation. suggested that other T cell populations can produce IL-17. Murine CD8+ T cells produce IL-17 in response to conditioned medium INTRODUCTION from dendritic cells (DCs) activated with Klebsiella pneumoniae (Happel et al., 2003). Furthermore, unconventional T cells can Th17 cells, CD4+ T cells that secrete interleukin-17 (IL-17), play also secrete IL-17 during infection (Lockhart et al., 2006; Shibata a critical role in inflammatory pathology in organ-specific autoim- et al., 2007). NKT cells, which express IL-23R, secrete IL-17 after mune diseases, including multiple sclerosis (MS) and rheuma- stimulation with anti-CD3 and IL-23 (Rachitskaya et al., 2008). toid arthritis, a role hitherto attributed to Th1 cells (Harrington gd T cells provide an early source of IFN-g in mucosal tissues et al., 2005; Koenders et al., 2005; Langrish et al., 2005; Park (Lochner et al., 2008; Ivanov et al., 2006) and are pathogenic in et al., 2005). Furthermore, Th17 cells play a role in immunity to EAE, collagen-induced arthritis (CIA), and colitis models in infection, for which they are generated through induction of mice (Jensen et al., 2008; Nanno et al., 2008; Roark et al., innate cytokines via activation of Toll-like receptors (TLRs) and 2007; Spahn et al., 1999). Their pathogenic role in autoimmunity NOD-like receptors (NLRs) (Higgins et al., 2006; Veldhoen was initially thought to be mediated through IFN-g production. et al., 2006). IL-6 and transforming growth factor (TGF)-b can However, recent studies have shown that gd T cells can secrete induce the differentiation of naive murine T cells into Th17 cells IL-17 during infection with Mycobacterium tuberculosis (MTB) or (Bettelli et al., 2006; Mangan et al., 2006; Veldhoen et al., E. coli, in which they promote neutrophil recruitment (Lockhart 2006). IL-21 is also a key cytokine in the differentiation of murine et al., 2006; Shibata et al., 2007). Interestingly, it has recently and human CD4+ Th17 cells, in which it appears to act in an been shown that IL-17 production by gd T cells is independent autocrine loop involving TGF-b or IL-23 (Korn et al., 2007; of IL-6 (Lochner et al., 2008), an essential cytokine for driving Nurieva et al., 2007; Yang et al., 2008; Zhou et al., 2007). the development of Th17 cells.

Immunity 31, 331–341, August 21, 2009 ª2009 Elsevier Inc. 331 Immunity gd T Cells Are an Important Source of Innate IL-17

A B Figure 1. IL-1b and IL-23 Promote IL-17 Production by CD3+ but Not by CD4+ or 150 No anti-CD3 CD8+ T Cells in the Absence of TCR Engage- 12.6 ment C57BL/6 Medium (A) Spleen cells from C57BL/6 or Il1r1À/À mice 100 -/- Il1r1 were stimulated with IL-1b and/or IL-23 with and without anti-CD3 for 3 days. IL-17 concentrations 50 in supernatants were quantiﬁed by ELISA. **p < IL-17 (pg/ml) 0.01 and ***p < 0.001, versus C57BL/6. ** (B) Spleen cells from C57BL/6 mice were stimu- 0 11.3 lated as indicated without anti-CD3 for 3 days and TGF- +IL-6 stained for intracellular IL-17 and for surface CD3. 1000 + Anti-CD3 IL-17 Numbers refer to percentages of CD3+IL-17+ cells. (C) Spleen cells, CD4+ T cells, or CD4+CD62LhiC- C57BL/6 750 D44lo T cells from C57BL/6 mice were stimulated Il1r1-/- for 3 days as indicated. IL-17 concentrations 500 were quantiﬁed by ELISA. **p < 0.01 and ***p < 30.4 0.001, versus medium control. IL-17 (pg/ml) IL-17 250 ** IL-1+IL-23 (D) Spleen cells from C57BL/6 mice were cultured *** *** for 3 days with medium only or IL-1b and IL-23, 0 and intracellular IL-17 was determined in CD3, TGF- - +-+ --- CD4, or CD8 cells. Numbers refer to percentages - IL-6 -- ++ -- of IL-17+ cells. Results for ELISA data are IL-1 -- -- - ++ CD3 means ± SE of triplicate culture and representative IL-23 -- -+- - + D of three experiments. All error bars represent C Medium IL-1+ IL-23 mean ± SD. No anti-CD3 + anti-CD28 100 *** 6.3 27.0 Spleen + 75 CD4+ CD3 CD4+CD62LhiCD44lo kine combinations induced IL-17 produc- 50 tion by murine spleen cells after stimula-

IL-17 (pg/ml) 25 tion with anti-CD3. In contrast, IL-6 and 9.8 8.4 TGF-b failed to induce IL-17 production 0 CD8+ IL-17 in the absence of TCR activation 600 + Anti-CD3 + anti-CD28 *** (Figure 1A), whereas IL-17 was produced 500 Spleen by spleen cells stimulated with IL-23 and + 400 CD4 IL-1b in the absence of anti-CD3. Further- CD4+CD62LhiCD44lo more, in the presence of anti-CD3, IL-17 300 3.1 3.8 ** CD4+ production in response to IL-23 and 200

IL-17 (pg/ml) IL-1b, but also in response to IL-6 and 100 TGF-b, was substantially reduced in 0 spleen cells from Il1r1À/À mice, demon- Medium IL-1 IL-23 IL-1+IL-23 Fsc strating the critical role of IL-1 signaling in IL-17 production. IL-17 protein production was assessed after 72 hr, but was In this study, we address the hypothesis that gd T cells may detectable at low amounts after 48 hr (Sutton et al., 2006) and secrete innate IL-17 in response to IL-1 and IL-23. We found was maximal in the region of 10 ng/ml of IL-1b and IL-23 (data that gd T cells expressed IL-23R and that stimulation with IL-1 not shown), a concentration that was used for subsequent exper- and IL-23 promoted RORgt, IL-17A, IL-17F, IL-21, and IL-22 iments. Intracellular cytokine staining (ICS) conﬁrmed that IL-1b expression by gd T cells in the absence of T cell receptor (TCR) and IL-23, but not IL-6 and TGF-b, induced IL-17 production by engagement. Furthermore, IL-17-secreting gd T cells were CD3+ T cells in the absence of anti-CD3 (Figure 1B). When acti- present at high frequency in the central nervous system (CNS) vated with anti-CD3 and anti-CD28, IL-1b and IL-23 failed to stim- of mice with EAE, in which they functioned with CD4+ Th17 cells ulate IL-17 production by naive (CD4+CD62LhiCD44lo) T cells but to mediate autoimmune pathology. did promote IL-17 production by total CD4+ T cells (Figure 1C). Furthermore, ICS demonstrated that, in the absence of anti- RESULTS CD3, IL-1b and IL-23 induced IL-17 secretion by CD3+ T cells, but not by CD4+ or CD8+ T cells (Figure 1D). These ﬁndings IL-1b and IL-23 Promote IL-17 Production by demonstrate that unconventional CD3+CD4ÀCD8À T cells, but Unconventional T Cells in the Absence of TCR Activation not conventional CD4+ or CD8+ T cells, secrete IL-17 in response Consistent with previous reports on the roles of IL-6 with TGF-b to IL-1b and IL-23 in the absence of TCR engagement. and IL-1b with IL-23 in promoting IL-17 production by naive and A current view on the generation of IL-17-producing T cells memory T cells, respectively, we found that both of these cyto- in vivo is that innate IL-6 and TGF-b promote development,

332 Immunity 31, 331–341, August 21, 2009 ª2009 Elsevier Inc. Immunity gd T Cells Are an Important Source of Innate IL-17

A IL-1b and IL-23, in the absence of stimulation with anti-CD3 5000 (Figure 3A). Furthermore, IL-17 was barely detectable in spleen 4000 B Medium À/À MTB 10 g/ml cells from mice defective in gd TCR (Tcrd ), but was produced 3000 3000 MTB 50 g/ml by spleen cells from wild-type (WT) mice stimulated with IL-1b 2000 and IL-23 in the absence of anti-CD3 (Figure 3B). Conversely,

IL-23 (pg/ml) 2000 1000 puriﬁed gd T cells, but not spleen cells depleted of gd T cells, 0 1000 secreted IL-17 in response to IL-1b and IL-23 (Figure 3C). ICS

2500 (pg/ml) Cytokine conﬁrmed that a substantial proportion of gd T cells secreted 2000 0 IL-23 IL-1 IL-12p70 IL-17 in response to IL-1b and IL-23 in the absence of anti- 1500 CD3 (Figure 3D). In contrast, we failed to detect IL-17 production 1000 by NK or NKT cells after stimulation of spleen cells with IL-1b and

IL-1 (pg/ml) C 500 1500 IL-23 (not shown). 0 Medium MTB We next used real-time PCR to confirm the effect of IL-1b and 70 1000 IL-23 on expression of IL-17 and other cytokines, receptors and 60 * transcription factors associated with Th17 cells. Stimulation 50 500 40 IL-17 (pg/ml) of purified gd T cells with IL-23 induced expression of IL-17A, 30 0 ** IL-17F, and IL-22 mRNA, which was enhanced by addition of IL-6 (ng/ml) 20 T cells + - + + + + 10 DC - + + + + + IL-1b, and this was significantly (p < 0.001) higher than that IL-1Ra - - - + - + 0 + anti-IL-23R - - - - + + induced in gd T cell-depleted CD3 T cells (Figure 3E). IL-21 Med LPS PGN MDP LPS MDP +PGN +PGN plays a role in the development of murine CD4+ Th17 cells (Korn et al., 2007; Nurieva et al., 2007; Zhou et al., 2007). We Figure 2. Killed M. tuberculosis or Purified TLR and NLR Agonists found that IL-1b and IL-23, but neither cytokine alone, promoted Induce Innate IL-17 Production by T Cells by Stimulating DC IL-23 substantial IL-21 expression in gd T cells, but not in CD3+ T cells and IL-1b Production depleted of gd T cells (Figure 3E). The transcription factor RORgt (A and B) Murine bone marrow-derived DCs were incubated with LPS, MDP, + PGN, LPS + PGN, MDP + PGN or with medium alone (A), or with killed MTB is associated with the development of CD4 Th17 cells (Ivanov (B). Supernatants were removed after 24 hr and IL-23, IL-1b, IL-12p70, and et al., 2006) and is also expressed by NKT cells (Rachitskaya IL-6 concentrations quantified by ELISA. et al., 2008). We demonstrate constitutive RORgt expression in (C) DCs were stimulated with MTB or medium only for 24 hr, washed, and gd T cells, which was enhanced by IL-23 (Figure 3E). It has previ- + added to CD3 T cells with IL-1Ra, anti-IL-23R, both, or a control antibody ously been reported that IL-23R is expressed on NK cells and (not shown). After 3 days, IL-17 was quantified in the supernatants by ELISA. memory T cells (Parham et al., 2002). We found that gd T cells ex- *p < 0.05 and **p < 0.01, versus MTB alone. Data are means for triplicate pressed high constitutive levels of IL-23R, which were enhanced cultures and are representative of two experiments. All error bars represent mean ± SD. by stimulation with IL-23 alone or in combination with IL-1b (Figure 3E). Cell-surface staining and flow cytometry analysis confirmed that IL-23 promoted IL-23R expression on gd T cells whereas IL-1 and IL-23 enhance the expansion of Th17 cells, and (Figure 3F). We also examined the capacity for IL-1 and IL-23 that these innate cytokines are induced in response to pathogen- to stimulate IL-17 production in vivo. We found substantial associated molecular patterns (PAMPs), including TLR and NLR IL-17A and IL-17F gene expression in the draining lymph nodes agonists. Here, we examined the role of TLR and NLR-induced 4 hr after injection of IL-1 and IL-23 into the footpad (Figure S2A). IL-1 and IL-23 production by DCs in promoting innate IL-17 Consistent with a previous report (Jensen et al., 2008), ICS production by T cells. Stimulation of DC with TLR and NLR revealed a relatively high frequency of IL-17-producing gd agonists (Figure 2A) or with killed MTB (Figure 2B) induced T cells in lymph nodes of naive mice. We found that this was high concentrations of IL-1b, IL-23, and IL-6, but interestingly enhanced after injection of IL-1 and IL-23 (Figure S2B). These + no IL-12p70. Furthermore, in the absence of anti-CD3, CD3 findings suggest that IL-1 and IL-23 promote rapid IL-17 gene T cells cultured with DC that had been simulated with MTB expression and protein production in vitro and in vivo and that (Figure 2C) or TLR and NLR agonists (Figure S1 available online), gd T cells are a source of innate IL-17, but do not rule out other induced IL-17 production, which was blocked by IL-1 receptor innate sources of IL-17. antagonist (IL-1Ra) and partially blocked by anti-IL-23R. These findings demonstrate that PAMP-activated IL-1b and IL-23 A High Frequency of IL-17-Secreting gd T Cells Infiltrate promote innate IL-17 production by T cells in the absence of the CNS of Mice with EAE TCR activation. Having shown that gd T cells produce IL-17 in response to stimulation with IL-1b and IL-23, we examined the possibility that IL-1b and IL-23 Induce IL-17 and IL-21 Expression these cells are a source of IL-17 during development of EAE, in in gd T Cells in the Absence of TCR Engagement which IL-17 has been shown to play a key role in inflammatory Having demonstrated that CD3+ but not CD4+ or CD8+ T cells pathology. We found substantial infiltration of IL-17-secreting secrete IL-17 in response to IL-1b and IL-23, in the absence of gd T cells into the brain during the course of EAE (Figure 4A). stimulation with anti-CD3 or anti-CD28, we sought to identify gd T cells were detectable on day 7 after induction, immediately the T cell population responsible for the innate IL-17 production. prior to onset of clinical signs of EAE, and peaked at day 10–14 We found that depletion of gd T cells significantly (p < 0.001) and a very high proportion (50%–80%) of these cells secreted IL- reduced the production of IL-17 by CD3+ T cells in response to 17 (Figures 4A and 4C). In contrast, the frequency of gd T cells

Immunity 31, 331–341, August 21, 2009 ª2009 Elsevier Inc. 333 Immunity gd T Cells Are an Important Source of Innate IL-17

Figure 3. IL-1b and IL-23 Stimulate IL-17A, A E +++ +++ 300 5000 *** 15000 *** IL-17F, IL-21, IL-22, and IL-23R Expression CD3+ by gd T Cells + 4000 -depleted CD3 + + 200 10000 +++ (A–C) CD3 T cells and CD3 T cells depleted of gd 3000 ** ++ T cells (A), spleen cells from C57BL/6 and TcrdÀ/À 2000 ** (fold increase) 100 increase) (fold 5000 mice (B), puriﬁed gd T cells, and gd-depleted

IL-17 (pg/ml) IL-17 *** 1000 spleen cells from C57BL/6 mice (C) were stimu-

*** Il17f Il17a *** 0 0 0 lated for 3 days as indicated. IL-17 concentrations - IL-1 IL-23 IL-1+ in supernatants were quantiﬁed by ELISA. ***p < IL-23 + 5000 400 *** 0.001 versus WT or nondepleted. B +++ + *** 300 *** (D) Spleen cells from C57BL/6 mice were cultured 4000 300 C57BL/6 *** with IL-1b and IL-23 or medium alone for 3 days, Tcrd-/- 3000 +++ *** 200 + and intracellular staining was performed for IL-17 200 *** 2000 gd (fold increase) (fold and surface TCR. Numbers refer to the (fold increase) (fold 100 1000 percentage of positive cells.

100 Il22 *** Il23r IL-17 (pg/ml) *** *** gd + gd 0 0 ** (E) -depleted CD3 T cells or puriﬁed T cells *** were stimulated with medium only, IL-1, IL-23, or 0 - IL-1 IL-23 IL-1+ + +++ *** IL-1 and IL-23 both. IL-17A, IL-17F, IL-21, IL-22, 20000 125 IL-23 *** RORgt, and IL-23R mRNA expression was evalu- *** 100 ated by real-time PCR normalized to 18 s rRNA C 300 + 75 -depleted *** and relative to gd-depleted T cells stimulated with 10000 medium only. *p < 0.05, **p < 0.01, and ***p < 200 50 ++ (fold increase) (fold (fold increase) (fold *** 0.001 versus medium control; +p < 0.05, 2+p< 25 +++ Il21 Rorc *** 0.01, and p < 0.001, versus gd-depleted T cells. 100 * IL-17 (pg/ml) IL-17 0 0 (F) Spleen cells were stimulated with IL-23 or ---- IL-1 + +++ IL-1 + ---- +++ medium for 24 hr and were then analyzed for 0 *** IL-23 --- - ++++ IL-23 --- - ++++ - IL-1 IL-23 IL-1+ IL-23R expression. Dot plots are shown for staining IL-23 -depleted T cells -depleted T cells with anti-IL-23R and an isotype control antibody. CD3 CD3 Numbers on the dot plots indicate the percentage of gated gd T cells. All data are representative of at D Medium F IL-1+IL-23 IL-23 Medium IL-23 least three independent experiments. All error bars represent mean ± SD. 12.8 31.4 0.4 1.6 5.5 IL-17 IL-23R Isotype

secreting IFN-g was significantly lower, and there was only IL-17-Secreting gd T Cells Cooperate a small increase in their numbers during the course of EAE. with CD4+ T Cells in the Development of EAE Conversely, the frequency of both CD4+IFN-g+ and CD4+IL- We next examined the role of gd T cells in the development of the 17+ T cells was enhanced during the development of EAE clinical symptoms of EAE. We found that EAE was less severe in (Figure 4B). TcrdÀ/À mice; the onset of disease was delayed and the clinical gd T cells expressing Vg4+ are expanded in CIA (Roark et al., scores were reduced (Figure 5A). This was associated with 2007). We found that more than 70% of gd T cells in the brain significantly (p < 0.01 – p < 0.001) lower MOG-specific IL-17 at the acute phase of EAE were Vg4+ and that more than 50% production by lymph nodes of TcrdÀ/À mice (Figure 5B), sug- of these secreted IL-17 (Figure 4D). We also detected Vg5+ cells gesting that gd T cells are required for antigen-specific and Vd6.3/2+ cells, but the majority of these were not IL-17 responses by conventional T cells during development of EAE. secreting. In the spleen, the overall frequency of IL-17-secreting Furthermore, total IL-17A and IL-17F gene expression in the gd T cells was lower but over 90% of these were Vg4+ spleen was significantly (p < 0.05) lower in TcrdÀ/À mice (Figure S3). compared with WT mice with EAE (Figure 5C), suggesting that We also found elevated IL-17 mRNA expression in the CNS gd T cells make a major contribution to IL-17 production in vivo during development of EAE (Figure 4E). In addition, IL-1b and during development of EAE. IL-23p19 mRNA was enhanced in the spleen 1–7 days after We next examined the role of gd T cells activated by IL-1b and immunization (data not shown), but was also enhanced in the IL-23 in the induction of EAE. Il1r1À/À mice were substantially CNS at the onset of clinical symptoms (Figure 4E), which coin- more resistant to EAE than wild-type mice (Figure 5D). Transfer cided with the detection of IL-17-secreting gd T cells in the brain. of CD3+ T cells from naive C57BL/6 mice to Il1r1À/À mice before These findings demonstrate that gd T cells are an important immunization with MOG and complete Freund’s adjuvant (CFA) source of IL-17 in the CNS during development of EAE and, enhanced susceptibility to EAE, but depletion of gd T cells together with the data presented in Figure 3, suggest that IL-1 from the transferred CD3+ T cells reduced their capacity to and IL-23 may be an important stimulus for early production of confer susceptibility to EAE. Transfer of autoantigen-specific innate IL-17 during development of EAE. Th17 cells has been shown to transfer EAE to naive mice

334 Immunity 31, 331–341, August 21, 2009 ª2009 Elsevier Inc. Immunity gd T Cells Are an Important Source of Innate IL-17

A B Figure 4. A High Frequency of IL-17-Secreting gd T Naive EAE Naive EAE Cells Infiltrate the CNS of Mice at the Onset of EAE, 31 0.2 51 2.8 33 15 43 27 which Correlates with IL-1b and IL-23p19 mRNA CD3 Expression CD3 EAE was induced in C57BL/6 mice. Brain mononuclear cells from naive mice and mice 7, 10, 14, and 21 days after induction of EAE, were stained for intracellular IL-17 or IFN-g and CD4 surface CD45, CD3, and gd TCR or CD4. Representative dot plots of IL-17 or IFN-g-secreting gd T cells (A) or IL-17 or + 8.3 83 0.1 17 IFN-g-secreting CD4 T cells (B) on day 10 are shown. (C) shows the mean percentage and absolute number of IL-17 or IFN-g-secreting gd T cells for five control (Cont.) mice and five mice at 7, 10, 14, and 21 days after induction of EAE; *p < 0.05, **p < 0.01, ***p < 0.001 versus control. As shown in (D), brain mononuclear cells from mice 10 days after induc- IL-17 CD4 IL-17 tion of EAE were stained for intracellular IL-17 and for surface 10.0 15 16 22 gd TCR, Vg4, Vg5, and Vd6.3/2. Numbers refer to percentages of total gd T cells. As shown in (E), IL-17A, IL-1b, and IL-23p19 mRNA expression was evaluated in the brain (cortex) by real- time PCR normalized to 18 s rRNA and relative to naive control mice. **p < 0.01, versus control. Dot plots are representative of five individual mice per group. Data are representative of three IFN- IFN- experiments. All error bars represent mean ± SD. C

90 + + *** 1000 + + 80 IFN- ** IFN- + + * 70 +IL-17+ IL-17 ** 750 *** 60 gd T Cells Activated with IL-1 and IL-23 50 500 + 40 Promote IL-17 Production by CD4 T Cells 30 % T cells % T * Our findings suggest that gd T cells, as well as 20 * ** 250 10 * directly contributing to pathogenic cytokine 0 brain / number Absolute 0 Cont. d7 d10 d14 d21 Cont. d7 d10 d14 d21 production, may also augment IL-17 production by autoantigen-specific Th17 cells in vivo. We D sought to provide further evidence for this hypoth- 44 6 51 2 21 26 esis. Consistent with the data in Figure 3, we found that spleen cells from TcrdÀ/À mice failed to IL-17 44 26 34 secrete IL-17 in response to IL-1b and IL-23, in the absence of anti-CD3 (Figure 6A). Addition of anti-CD3 (required to activate IL-17 production V4 V5 Vδ6.3/2 by Th17 cells) substantially augmented IL-17 E production induced by IL-1b and IL-23 or either 20 40 ** 5 cytokine alone. However, this was also signifi- ** 16 4 cantly (p < 0.001) reduced in spleen cells from 30 À/À 12 3 Tcrd mice (Figure 6A), suggesting that gd 20 T cells can augment IL-17 production from ab 8 2 (fold increase) (fold (fold increase) (fold increase) (fold 10 T cells. This was confirmed by the demonstration 4 1 + Il1b Il23a Il17a that IL-17 production by CD4 T cells stimulated 0 0 0 Cont.12h d7 d14 d21 Cont.12h d7 d14 d21 Cont.12h d7 d14 d21 with anti-CD3 and anti-CD28 was significantly (p < 0.001) enhanced by addition of gd T cells that had been prestimulated with IL-1b and IL-23 (and washed to remove IL-1b and IL-23), but not with (Langrish et al., 2005). We found that lymph node and spleen gd T cells stimulated with medium only (Figure 6B). Addition of cells from mice with EAE stimulated with MOG in the presence DCs to the cultures enhanced the ability of IL-1b- and IL-23-acti- of IL-1 and IL-23 induced EAE after transfer to recipient mice vated gd T cells to promote IL-17 production by CD4+ T cells (Figure 5E). These cultured cells included IL-17-secreting gd stimulated with anti-CD3 and anti-CD28 (Figure 6C). These find- T cells as well as CD4+ T cells, but not NK cells (data not shown). ings suggest that IL-1b- and IL-23-activated gd T cells promote CD4+ and CD8+ with gd T cells purified from the cultured cells IL-17 production by ab T cells, either by acting directly on the also induced mild EAE, but purified CD4+, CD4+ and CD8+,or CD4+ T cells and/or indirectly by promoting cytokine production gd T cells alone failed to transfer EAE (Figures 5E and 5F). Collec- from DCs. tively, our data suggest that IL-17-producing gd T cells, as well In order to examine the relative role of soluble factors as CD4+ T cells, infiltrate the CNS of mice with EAE and that produced by gd T cells activated with IL-1b and IL-23 in both of these cell types are required for optimal development promoting IL-17 production by CD4+ T cells, we stimulated gd of EAE. T cells with IL-1b and IL-23 overnight, then washed and

Immunity 31, 331–341, August 21, 2009 ª2009 Elsevier Inc. 335 Immunity gd T Cells Are an Important Source of Innate IL-17

A D C57BL/6 Figure 5. gd T Cells Cooperate with Conven- -/- + tional T Cells in the Development of EAE * Il1r1 + WT CD3 4 C57BL/6 5 Il1r1-/- + WT -depleted CD3+ (A and B) EAE was induced in C57BL/6 or TcrdÀ/À Tcrd-/- -/- mice. * 4 Il1r1 3 * (A) Mice were observed daily for clinical symptoms 3 of EAE. *p < 0.05 versus TcrdÀ/À. 2 (B) Mice were sacriﬁced and lymph node cells 2 stimulated for 3 days with MOG peptide, medium, Clinical score Clinical 1 scoreClinical 1 or PMA and anti-CD3. IL-17 concentrations in supernatants were determined by ELISA. **p < 0 0 0.01 and ***p < 0.001 versus C57BL/6. 0 5 10 15 20 25 10 20 30 (C) IL-17A and IL-17F mRNA expression was eval- Time post immunization (days) Time post immunization (days) uated in spleen tissue by real-time PCR normalized B E to 18 s rRNA and relative to naive control mice. *p < MOG Unseparated 0.05 versus WT with EAE. 600 Medium 2.5 MOG + IL-1 + IL-23 Unseparated (D) EAE was induced in C57BL/6 or Il1r1À/À mice. 500 MOG 12.5 + ** MOG + IL-1 + IL-23 In addition, CD3+ T cells or CD3+ T cells depleted 400 MOG 25 2.0 + MOG + IL-1 + IL-23 CD4 of gd T cells from WT mice were transferred into 300 MOG 50 150 1.5 À/À PMA/aCD3 Il1r1 mice 24 hr before induction of EAE. Mice were observed daily for clinical signs of EAE; n = 5. 100 1.0 (E and F) Spleen and lymph node cells from mice IL-17 (pg/ml) 50 Score Clinical 0.5 with EAE were stimulated in vitro with MOG or ** ** *** MOG + IL-1 + IL-23. After 3 days, unseparated 0 0.0 cells, CD4+ T cells, or gd T cells puriﬁed from the -/- 0 10 20 30 C57BL/6 Tcrd cultures (E) or CD4+ and CD8+ or CD4+ and CD8+ Time post transfer (days) plus gd T cells (F) were transferred to naive recipient mice. Mice were observed for symptoms of EAE. C F Data are representative of two to three independent 2.5 + + + 125 10 MOG + IL-1 + IL-23 CD4 + CD8 + experiments. All error bars represent mean ± SD. + + 2.0 MOG + IL-1 + IL-23 CD4 + CD8 100

75 1.5 5 50 1.0 (fold increase) (fold

(fold increase) * 25 Clinical Score 0.5 IL-1b and IL-23 was inhibited by addition Il17f Il17a * 0 0 0.0 of anti-IL-17 or anti-IL-21; the greatest -/- -/- WT WT Tcrd WT WT Tcrd 0 10 20 30 40 50 60 70 inhibition of IL-17 expression was Cont EAE EAE Cont EAE EAE Time post transfer (days) observed when both antibodies were added to the cultures (Figure 6D). These recultured them for an additional 48 hr. Supernatants were findings suggest that gd T cell-derived IL-17 and IL-21 augment recovered at 72 hr and used for CD4+ T cell stimulation. Super- IL-17 production by Th17 cells. natants from gd T cells that had been stimulated with IL-1b and IL-23 induced IL-17A and IL-17F mRNA expression by CD4+ T Cells and DCs Express IL-17R CD4+ T cells (Figure 6D). These findings demonstrate that gd and Respond to IL-17 T cells activated with IL-1b and IL-23 can promote anti-CD3- In order to provide further evidence that IL-17 may amplify its induced IL-17 production by CD4+ T cells, at least in part through own production either by acting directly on CD4+ T cells or indi- the release of soluble factors. Given that binding of ligands to aryl rectly on DC, we examined expression of IL-17R and responsive- hydrocarbon receptors (AhRs) expressed on Th17 cells induces ness of these cells to IL-17. We first showed that CD4+ T cells IL-17 production (Veldhoen et al., 2009), we assessed the from naive mice express IL-17R mRNA and the expression possible role of AhR signaling in IL-1- and IL-23-induced IL-17 was stronger than that observed for lung or spleen (Figure 7A), production. We found that an AhR antagonist, CH-223191, did tissues previously shown to express IL-17R. We next used not suppress IL-17 production by spleen cells stimulated with IL-17 mRNA expression to examine autocrine production of IL-1 and IL-23 in the presence or absence of anti-CD3 IL-17 by CD4+ T cells and found that recombinant IL-17A (Figure S4). induced IL-17A and IL-17F mRNA expression in CD4+ T cells We have shown that IL-1b and IL-23 can induce IL-17 and (Figure 7B). We also examined the effect of IL-21 and found IL-21 expression in gd T cells, and it has already been reported that it did not induce IL-17 or amplify IL-17-induced IL-17 that IL-21 together with TGF-b or IL-23 can stimulate IL-17 expression by CD4+ T cells (not shown). However, IL-21 did production by CD4+ T cells (Korn et al., 2007; Nurieva et al., enhance IL-23-induced IL-17A and IL-17F expression by CD4+ 2007; Zhou et al., 2007). However, it is also possible that gd T cells (Figure 7C). T cell-derived IL-17 may promote further IL-17 production by DCs were also found to express IL-17R, which was enhanced Th17 cells. Therefore, we examined the roles of IL-21 and after stimulation with TLR agonists (Figure 7D). DCs secrete IL-17 in gd T cell-induced IL-17 production by CD4+ T cells. IL-23 after stimulation with low concentrations of IL-17 (Fig- We found the induction of IL-17A and IL-17F mRNA expression ure 7E), and this was inhibited by coincubation with IL-17RFc in CD4+ T cells by supernatants from gd T cells activated with (Figure S5). Furthermore, IL-17-induced IL-23 production was

336 Immunity 31, 331–341, August 21, 2009 ª2009 Elsevier Inc. Immunity gd T Cells Are an Important Source of Innate IL-17

with anti-IL-17R (Figure S6A). Moreover, anti-IL-17R treatment A 1200 D 15 C57BL/6 suppressed IL-17A mRNA expression in the spinal cord 1000 Tcrd-/- 800 10 (Figure S6B). The reduction in IL-17 was associated with slower 600 onset of EAE in these mice (Figure S6C). Conversely, we found

400 increase) (fold 5 that administration of IL-17 with CFA and MOG resulted in earlier

IL-17 (pg/ml) *** 200 onset of EAE and more pronounced clinical scores early in *** *** Il17f * *** 0 0 disease (Figure S6D). IL-1 -++-++-- IL-23 -- + ++-- + These data suggest that early IL-17 has a positive rather than αCD3 - - -+- ++ + 5.0 negative feedback role on IL-17 production, especially early in B +++ EAE. Collectively, our ﬁndings suggest that IL-1b and IL-23 1200 *** 1000 promote innate IL-17 production by gd T cells, which ampliﬁes 2.5 * + 800 * IL-17 production by CD4 T cells, either by interacting directly (fold increase) (fold + 600 ** with CD4 T cells or indirectly with DC to enhance production Il17a 400 *** of innate cytokines that drive the induction and expansion of IL-17 (pg/ml) *** 0.0 200 CD4+ T cells +++ + + Th17 cells. Thus, gd T cells, as well as making a substantial γδ (IL-1+IL-23) - + +++ + 0 contribution to the overall pool of IL-17 in vivo, may act as an CD4 T cells - - + +++ anti-IL-17 -- - + + γδ (medium) + --- + - anti-IL-21 - +- - + early source of IL-17 and IL-21 to amplify Th17 cell responses. γδ (IL-1β+IL-23) --+ - - + αCD3/αCD28 --- +++ DISCUSSION ### C 800 *** 700 600 The results of this study demonstrate that gd T cells are an impor- 500 tant source of innate IL-17 and that these cells constitutively 400 300 express IL-23R and also secrete IL-21 and IL-22. The present

IL-17 (pg/ml) IL-17 200 study also reveals a new mechanism of IL-23 and IL-1-induced *** 100 IL-17 production. We demonstrate that IL-23 in combination + 0 CD4 T cells + + + +++ with IL-1b promotes IL-17, IL-21, and IL-22 expression by gd γδ (medium) - --+ + - γδ (IL-1β+IL-23) --- + - + T cells in the absence of additional signals. Finally, we demon- DC + αCD3 --- +++ strate that gd T cells may act in an amplification loop to promote IL-17 production by Th17 cells. Figure 6. gd T Cells Activated with IL-1b and IL-23 Promote IL-17 The transcription factor RORgt, which has been associated + Production from CD4 T Cells with the development of CD4+ Th17 cells (Ivanov et al., 2006), À/À (A) Spleen cells from C57BL/6 and Tcrd mice were stimulated as indicated is also expressed by NKT cells (Rachitskaya et al., 2008) and for 3 days and IL-17 concentrations in supernatants quantified by ELISA. CD4+CD3À lymphoid tissue inducer (LTi)-like cells (Takatori ***p < 0.001 versus WT. (B and C) gd T cells were incubated with IL-1b and IL-23 or medium alone. After et al., 2009). We demonstrated constitutive expression of RORgt 3 days, cells were washed and added to purified CD4+ T cells with and without in gd T cells, which was enhanced by IL-23. Furthermore, IL-23 anti-CD3 and anti-CD28 (B) or anti-CD3 and DCs (C). After an additional induced IL-17A, IL-17F, and IL-22 expression, which was 3 days, supernatants were removed and IL-17 concentrations quantified by enhanced by IL-1b. Interestingly, we found that IL-21, which is +++ ELISA. ***p < 0.001 versus no anti-CD3, p < 0.001 gd (medium) versus gd induced in naive CD4+ T cells by stimulation with anti-CD3, (IL-1b + IL-23), ###p < 0.001, versus without DCs. anti-CD28, and IL-6 or IL-6 and TGF-b (Korn et al., 2007), was (D) Purified CD4+ T cells were stimulated with anti-CD3 and anti-CD28 and medium only or with supernatants from gd T cells stimulated with IL-1b + IL- also expressed at very large amounts in gd T cells, but not by + 23 (washed on day 1), with or without anti-IL-17, anti-IL-21, or both antibodies. gd T cell-depleted CD3 T cells, after stimulation with IL-1b IL-17A and IL-17F mRNA expression was evaluated by real-time PCR normal- and IL-23. ized to 18 s rRNA and relative to CD4+ T cells stimulated with medium only. In addition to the roles of IL-6 and TGF-b in promoting the *p < 0.05 and **p < 0.01, with versus without antibody. Data are representative differentiation of Th17 cells, there are firmly established roles of three independent experiments. All error bars represent mean ± SD. for IL-23 and IL-1 in promoting IL-17 production and mediating the development of autoimmune diseases. A major function of enhanced by costimulation with a TLR agonist (Figure 7F). More- IL-23 in autoimmunity and infection is considered to involve over, LPS-induced IL-12p40, IL-1b, and IL-6 production by DC promoting IL-17 production from memory CD4+ T cells (Higgins was substantially enhanced by IL-17 (Figure 7F). In addition, et al., 2006; van Beelen et al., 2007). However, studies in an EAE IL-17 induced TGF-b and CCL2 mRNA expression and substan- model, involving adoptive transfer of T cells into Il23aÀ/À mice, tially augmented LPS-induced IL-1b, IL-6, CCL2, and CXCL10 showed that IL-23 played a critical role in the development but mRNA expression (Figure 7G). Addition of IL-17RFc reduced not in the effector function of encephalitogenic T cells (Thakker expression to that induced with LPS alone, confirming that the et al., 2007). In an EAU model, neutralization of IL-23 reduced effect of IL-17 was mediated via IL-17R and not due to a contam- IL-17 production and attenuated EAU when administered imme- inant in the IL-17 preparation. diately before and after induction of disease, but not at the We next examined an autocrine role for IL-17 in vivo by admin- effector stage of the disease (Luger et al., 2008). These findings istering anti-IL-17R or IL-17 during the induction of EAE. We are consistent with the suggestions that IL-23 may be central to found a significant (p < 0.05) reduction in antigen-specific differentiation rather than expansion or survival of pathogenic IL-17 production in lymph nodes, but not spleen, of mice treated Th17 cells (McGeachy and Cua, 2008; Wilson et al., 2007).

Immunity 31, 331–341, August 21, 2009 ª2009 Elsevier Inc. 337 Immunity gd T Cells Are an Important Source of Innate IL-17

G Figure 7. IL-17A Induces IL-17A and IL-17F A E + 25 800 25 Production from CD4 T Cells and Enhances 700 TLR-Induced IL-1b and IL-23 Production 20 600 20 from DCs 15 500 15 (A) IL-17R mRNA expression was evaluated in CD4+ 400 10 10 T cells, lung, spleen, and control tissue (Cont; liver) 300 increase) (fold (fold increase)

IL-23 (pg/ml) 200 from naive mice by real-time PCR normalized to 5 5

Il17r 100 18 s rRNA and relative to control tissue. Tgfb1 0 0 0 (B) Puriﬁed CD4+ T cells from C57BL/6 mice were Cont. Lung Spl. CD4 - 0.1 1 10 stimulated with anti-CD3 in the presence of IL-17 B IL-17 (ng/ml) 1750 12 or medium only (Med), and IL-17A and IL-17F Il17a mRNA 1500 10 Il17f mRNA mRNA expression was evaluated by real-time F 1250 5000 PCR normalized to 18 s rRNA. 8 1000 (C) Puriﬁed CD4+ T cells from C57BL/6 mice were 6 4000 750

(fold increase) stimulated with anti-CD3 and IL-21, IL-23, or both 500 4 3000

Fold increase cytokines or with medium only (Med). IL-17A and 2 Il1b 250 2000 0 IL-17F mRNA expression was evaluated by real- 0 IL-23 (pg/ml) time PCR normalized to 18 s rRNA and expressed Med IL-17 1000 6000 relative to medium only. C 20 0 5000 (D) BMDC were stimulated with medium (Med) 75 4000 only, LPS, or CpG, and IL-17R mRNA expression 15 3000 was evaluated by real-time PCR normalized to 2000 18 s rRNA and relative to a control (Cont.; liver). 50 increase) (fold 10 (E) BMDC were stimulated with IL-17 or medium

Il6 1000 (fold increase) (fold only. After 24 hr, IL-23 production in supernatants 5 0 25 was quantiﬁed by ELISA. Il17a IL-1 (pg/ml) 0 120 (F) DCs were stimulated with LPS, IL-17, LPS + 175 0 100 IL-17, or medium alone. After 24 hr, supernatants 150 were removed and IL-23, IL-1b, IL-12p40, and 175 80 125 IL-6 concentrations were quantiﬁed by ELISA. 150 60 100 (G) DCs were stimulated with LPS alone or with 125 (fold increase) 40 75 IL-17 in the presence or absence of IL-17RFc. After (fold increase) (fold 100 20 50 Ccl2 24 hr, TGF-b, IL-1b, IL-6, CCL2, and CXCL10 75 Il17f 25 0 mRNA was evaluated by real-time PCR normalized 0 50

IL-12p40 (ng/ml) to 18 s rRNA and relative to DC stimulated with Med IL-21 IL-23 IL-21 25 800 medium only. Data are representative of at least +IL-23 0 700 D 600 two independent experiments. All error bars repre- 30 15 500 sent mean ± SD. 400 20 10 increase) (fold 300 200 100 Cxcl10

(fold increase) (fold 10 5 0 IL-6 (ng/ml) IL-17 - 1110100 --110100 101001 10 100 (Takatori et al., 2009). It has also been re- Il17r LPS - --- + +++--+-- ++ ported that gd T cells and other non CD4+ 0 0 IL-17RFc ------+++++ ++ + Cont. DC DC DC IL-17 - + - + or CD8 T cells are a major source of IL-17 Med LPS CpG LPS - - + + during infection with M. tuberculosis (Lockhart et al., 2006). Vd1+ cells have Furthermore, studies in a colitis model have suggested that IL-23 been shown to accumulate at the site of E. coli infection, where can promote innate IL-17 production, but did not define the cell they promote neutrophil recruitment (Shibata et al., 2007). More- population involved (Hue et al., 2006). It has also been reported over, there is evidence that gd T cells may have a pathogenic role that stimulation of peritoneal exudate cells with IL-23, but not by in CIA (Roark et al., 2007). Our study demonstrates that gd T cells IL-6 and TGF-b, induced IL-17 production by gd T cells (Shibata infiltrate the CNS during the development of EAE, and a very high et al., 2007). Our study demonstrates a requirement for IL-1 and frequency of these cells secretes IL-17. We detected Vg4+,Vg5+, shows that IL-23 and IL-1b promote innate IL-17 production from and Vd6.3/2+ cells, but consistent with the CIA studies (Roark gd T cells and provides an alternative mechanism whereby IL-23 et al., 2007), the dominant IL-17-secreting gd T cells were and IL-1b may contribute to host defense and autoimmune Vg4+. A much smaller number of gd T cells in the CNS of mice inflammation. with EAE secreted IFN-g. In contrast, a similar frequency of It has previously been reported that unconventional T cells, IL-17 and IFN-g-secreting CD4+ T cells was detected in the especially NKT cells, express the IL-23R and secrete IL-17 CNS at the onset of EAE, and although the absolute numbers during infection or inflammation or when activated in vitro with are greater, a lower percentage of CD4+ than gd T cells secreted anti-CD3 and IL-23 or with the synthetic ligand a-galactosylcer- IL-17. A high frequency of IL-17-secreting gd T cells has also amide (Lockhart et al., 2006; Rachitskaya et al., 2008; Shibata been reported in draining lymph nodes 0–4 days after immuniza- et al., 2007). Furthermore, it has recently been demonstrated tion with MOG and CFA, which preceded the generation of that IL-23 can promote innate IL-17 and IL-22 production by MOG-specific CD4+ T cells (Jensen et al., 2008). Taken together LTi-like cells when stimulated in vitro or in vivo with zymosan with our findings, this suggests that IL-17-secreting gd T cells

338 Immunity 31, 331–341, August 21, 2009 ª2009 Elsevier Inc. Immunity gd T Cells Are an Important Source of Innate IL-17

may play a pathogenic role early in disease and may then influ- tion of DCs enhanced the ability of gd T cells activated with ence the generation of Th17 cells. Consistent with this, studies IL-1 and IL-23 to promote IL-17 production by CD4+ T cells. in humans have described an accumulation of gd T cells in acute We found that DCs also expressed IL-17R and secreted IL-23 MS lesions (Wucherpfennig et al., 1992) and clonal expansion of in responses to IL-17, which was enhanced by LPS and blocked gd T cells in cerebrospinal fluid of MS patients with recent-onset by addition of IL-17RFc. Furthermore, DCs expressed IL-1b, disease (Shimonkevitz et al., 1993). IL-6, TGF-b, CCL2, and CXCL10 in response to IL-17 alone or Evidence that the infiltrating IL-17-secreting gd T cells may in synergy with a TLR agonist. These findings demonstrate that have a pathogenic function in EAE is provided by our findings, gd T cell-derived IL-17 promotes further IL-17 production by which are consistent with previous reports (Jensen et al., 2008; Th17 cells, both directly by interacting with IL-17R on CD4+ Spahn et al., 1999) that clinical symptoms are attenuated in T cell and indirectly by enhancing the production of IL-23 and TcrdÀ/À mice. Furthermore, we found that depletion of gd other cytokines and chemokines from DCs that promote the T cells from CD3+ T cells from naive WT mice prior to transfer development, expansion, and recruitment of Th17 cells. to Il1r1À/À mice reduced their ability to confer susceptibility to Our study provides evidence that gd T cells, through IL-17 and EAE. In addition, we found that gd T cells were required for trans- IL-21 production, may help to mediate the effects of IL-23 and fer of EAE with cultured CD4+ and CD8+ T cells stimulated with IL-1b in promoting the development and expansion of Th17 cells. MOG, IL-1b, and IL-23. As well as contributing to the pool of Thus, analogous to the role of NK cells in amplifying the effects of IL-17 produced locally in the CNS during development of EAE, IL-12 on Th1 cells, gd T cells may act to amplify the induction of our data suggest that gd T cells may also play a cooperative Th17 cells. Our findings shed new light on the potential role of gd role with conventional autoantigen-specific ab T cells in medi- T cells in host defense and inflammatory diseases, provide ating autoimmune disease. Further evidence for this hypothesis important new information on the pathogenic role of IL-23 and was provided by the observation that attenuation of EAE in IL-1b, and underline the importance of targeting these cytokines TcrdÀ/À mice correlated with a substantial reduction in total in the development of new therapeutic interventions against IL-17 mRNA expression, but also in MOG-specific IL-17, which autoimmune diseases. is primarily produced by CD4+ T cells. Our study revealed that cytokines produced by gd T cells in EXPERIMENTAL PROCEDURES response to stimulation with IL-1 and IL-23 promote IL-17 expression by CD4+ T cells. Supernatants from gd T cells acti- Mice vated with IL-1 and IL-23 (washed and recultured for removal C57BL/6 (Harlan UK), Il1r1À/À, and TCR d mutant mice, defective in gd TCR À/À of IL-1 and IL-23) promoted IL-17A and IL-17F mRNA expression expression (Tcrd ) on C57BL/6 backgrounds (Jackson Labs), were bred under SPF conditions. Animal experiments and maintenance were approved by CD4+ T cells after costimulation with anti-CD3 and anti-CD28. by the university’s ethics committee and the Irish Department of Health. Furthermore, this enhancement of IL-17 mRNA expression was inhibited by coculture with anti-IL-21 and anti-IL-17. Because Induction and Assessment of EAE we also found that IL-21 was expressed by gd T cells after stim- C57BL/6, TcrdÀ/À, and Il1r1À/À mice were immunized subcutaneously (s.c.) ulation with IL-1 and IL-23, but not either cytokine alone, these with 150 mg of MOG peptide 35-55 (Cambridge BioSciences) emulsified in findings suggest that IL-21 may in part mediate the effect of gd CFA containing 4 mg/ml heat-killed MTB (Chondrex). Mice were injected intra- T cells on IL-17 production by CD4+ T cells. This conclusion peritoneally (i.p.) with 500 ng pertussis toxin (PT; Kaketsuken) on days 0 and 2. + + is consistent with the recent demonstration that IL-21 plays a In some experiments, purified CD3 T cells or CD3 T cells depleted of gd critical role in IL-17 production by naive CD4+ T cells (Korn T cells from naive mice were injected intravenously (i.v.) 24 hr prior to the induction of EAE. Alternatively, EAE was induced by transfer of spleen and et al., 2007; Nurieva et al., 2007; Zhou et al., 2007). The induction lymph node cells from mice with EAE (day 21) after culture for 3 days with either + of IL-17A and IL-17F mRNA expression in CD4 T cells by super- MOG or MOG, IL-1b, and IL-23 or by transfer of CD4+ or gd T cells purified from natants from gd T cells that had been stimulated with IL-1b and these cultures. Mice were observed for clinical signs of disease daily. Disease IL-23 was also inhibited by anti-IL-17. This was consistent with severity was recorded as follows: grade 0, normal; grade 1, limp tail; grade 2, our demonstration that CD4+ T cells expressed endogenous wobbly gait; grade 3, hind limb weakness; and grade 4, hind limb paralysis. IL-17R and responded to IL-17 by enhancing expression of IL- 17A and IL-17F mRNA. However, it has been reported that T Cell Sorting and Culture + + + hi lo IL-17-producing T cells are enhanced in the spleen or liver, but CD3 , CD4 , and CD4 CD62L CD44 T cells were purified from spleens of not the lymph node, of Il17raÀ/À mice and that IL-17F can C57BL/6 mice with mouse T cell subset column kits (R&D). gd T cells were + sorted by FACS (Dako Moflo) or MACS. Cells were cultured with IL-1b suppress IL-17A production by CD4 T cells (Nagata et al., (10 ng/ml), IL-23 (10 ng/ml), TGF-b (10 ng/ml), or IL-6 (20 ng/ml), in the absence 2008; Smith et al., 2008). We found that treatment with anti-IL- or presence of anti-CD3 (1 mg/ml) and anti-CD28 (10 mg/ml). After 24–72 hr, 17R early in EAE reduced IL-17 production and attenuated IL-17 concentrations in supernatants were determined by ELISA (R&D). disease, whereas administration of IL-17 with MOG and CFA enhanced the onset of EAE. Although an autocrine or paracrine Dendritic Cells role for IL-17 in modulating IL-17 production is still unclear, our Mouse bone-marrow derived immature DCs were generated as described findings suggest that early IL-17 has a positive rather than nega- (Higgins et al., 2006). DCs were cultured with MTB (10–50 mg/ml; Chondrex), tive feedback role for induction of IL-17, especially early in EAE. LPS (10–100 ng/ml; Alexis), CpG (20 mg/ml), muramyl dipeptide (20 mg/ml), peptidoglycan (25 mg/ml), or combinations thereof, with and without IL-1Ra This is consistent with our conclusions that early IL-17 from gd (10 mg/ml; Anakinra) and anti-IL-23R (10 mg/ml; R&D) or with IL-17 (0.1– T cells may promote the induction or activation of Th17 cells. 10 ng/ml; R&D) in the presence and absence of IL-17RFc (2 mg/ml; R&D). After Our study also reveals an additional mechanism whereby 24 hr, supernatants were tested for IL-1b, IL-12p70 (R&D), IL-23 (eBioscien- IL-17-producing gd T cells may amplify Th17 responses. Addi- ces), IL-6, and IL-12p40 (BD PharMingen) concentrations by ELISA.

Immunity 31, 331–341, August 21, 2009 ª2009 Elsevier Inc. 339 Immunity gd T Cells Are an Important Source of Innate IL-17

Intracellular Cytokine Staining Bettelli, E., Carrier, Y., Gao, W., Korn, T., Strom, T.B., Oukka, M., Weiner, H.L., Spleen cells (2 3 106 cells/ml) from C57BL/6 mice were incubated for 72 hr and Kuchroo, V.K. (2006). Reciprocal developmental pathways for the gener- with medium or cytokines and for the last 6 hr in the presence of brefeldin A ation of pathogenic effector TH17 and regulatory T cells. Nature 441, 235–238. (5–10 mg/ml). Alternatively, mononuclear cells were isolated by percoll density Chung, Y., Chang, S.H., Martinez, G.J., Yang, X.O., Nurieva, R., Kang, H.S., centrifugation from the brain of mice perfused with PBS and stimulated with Ma, L., Watowich, S.S., Jetten, A.M., Tian, Q., and Dong, C. (2009). Critical PMA (10 ng/ml) and ionomycin (1 mg/ml) with brefeldin A (5 mg/ml). Cells regulation of early Th17 cell differentiation by interleukin-1 signaling. Immunity were washed and blocked with Fcg blocker (BD PharMingen; 1 mg/ml) before 30, 576–587. extracellular staining for surface CD3, gd TCR, Vg4, Vg5, Vd6.3/2, CD4, or Cua, D.J., Sherlock, J., Chen, Y., Murphy, C.A., Joyce, B., Seymour, B., CD8. Cells were then fixed and permeabilized (Fix and Perm Cell Permeabili- Lucian, L., To, W., Kwan, S., Churakova, T., et al. (2003). Interleukin-23 rather zation Kit; Caltag) and stained intracellularly for IL-17 or IFN-g. Flow cytometric than interleukin-12 is the critical cytokine for autoimmune inflammation of the analysis was performed with either a BD FACS Caliber and Summit software or brain. Nature 421, 744–748. a Cyan ADP Flow Cytometer (Dako Cytomation) with FloJo software, with gating set on isotype controls or FMOs, respectively. Happel, K.I., Zheng, M., Young, E., Quinton, L.J., Lockhart, E., Ramsay, A.J., Shellito, J.E., Schurr, J.R., Bagby, G.J., Nelson, S., and Kolls, J.K. (2003). Cutting edge: Roles of Toll-like receptor 4 and IL-23 in IL-17 expression in Real-Time PCR response to Klebsiella pneumoniae infection. J. Immunol. 170, 4432–4436. gd T cells or gd-depleted T cells were purified from spleens of C57BL/6 mice and incubated for 48 hr with IL-1b (10 ng/ml), IL-23 (10 ng/ml), both, or medium Harrington, L.E., Hatton, R.D., Mangan, P.R., Turner, H., Murphy, T.L., Murphy, alone. CD4+ T cells were stimulated with supernatants from gd T cells activated K.M., and Weaver, C.T. (2005). Interleukin 17-producing CD4+ effector T cells with IL-1 and IL-23, in the presence or absence of anti-IL-17 and anti-IL-21 develop via a lineage distinct from the T helper type 1 and 2 lineages. Nat. Im- (10 mg/ml) or with recombinant IL-17 (10 ng/ml; R&D). DCs were stimulated munol. 6, 1123–1132. with medium only, LPS, or CpG. RNA was harvested from gd T cells, Higgins, S.C., Jarnicki, A.G., Lavelle, E.C., and Mills, K.H. (2006). TLR4 medi- gd-depleted T cells, CD4+ T cells, DCs, spinal cord, lung, liver, or spleen tissue ates vaccine-induced protective cellular immunity to Bordetella pertussis: with the Trizol (Invitrogen)/chloroform method; this process was followed by Role of IL-17-producing T cells. J. Immunol. 177, 7980–7989. reverse transcription into cDNA with a QuantiTect Reverse Transcription Kit Hue, S., Ahern, P., Buonocore, S., Kullberg, M.C., Cua, D.J., McKenzie, B.S., (QIAGEN). Real-time PCR for the detection of Il17a (Mm00439619_m1), Il17f Powrie, F., and Maloy, K.J. (2006). Interleukin-23 drives innate and T cell-medi- (Mm00521423_m1), Il23r (Mm00519942_m1), Rorc (Mm01261022_m1), Il21 ated intestinal inflammation. J. Exp. Med. 203, 2473–2483. (Mm00517640_m1), Il22 (Mm00444241_m1), Il17r (Mm00434214_m1), Tgfb1 Ivanov, I.I., McKenzie, B.S., Zhou, L., Tadokoro, C.E., Lepelley, A., Lafaille, (Mm00441724_m1), Ccl2 (Mm00441242_m1), Cxcl10 (Mm99999072_m1), J.J., Cua, D.J., and Littman, D.R. (2006). The orphan nuclear receptor ROR- Il23a (Mm00518984_m1), and Il1b (Mm00434228_m1) mRNA was performed gammat directs the differentiation program of proinflammatory IL-17+ T helper with predesigned Taqman gene expression assays (Applied Biosystems). cells. Cell 126, 1121–1133. Expression was normalized to 18s ribosomal RNA. Samples were assayed on an Applied Biosystems 7500 Fast Real Time PCR machine. Jensen, K.D., Su, X., Shin, S., Li, L., Youssef, S., Yamasaki, S., Steinman, L., Saito, T., Locksley, R.M., Davis, M.M., et al. (2008). Thymic selection determines Antigen-Specific IL-17 Production gammadelta T cell effector fate: Antigen-naive cells make interleukin-17 and 6 antigen-experienced cells make interferon gamma. Immunity 29, 90–100. Spleen and lymph node cells (1 3 10 cells/ml) were stimulated with MOG (12.5, 25 and 50 mg/ml) or PMA (Sigma; 20 ng/ml) and anti-CD3 (1 mg/ml), Koenders, M.I., Lubberts, E., Oppers-Walgreen, B., van den Bersselaar, L., and IL-17 concentrations in supernatants were determined by ELISA (R&D). Helsen, M.M., Kolls, J.K., Joosten, L.A., and van den Berg, W.B. (2005). Induc- tion of cartilage damage by overexpression of T cell interleukin-17A in exper- Statistical Analysis imental arthritis in mice deficient in interleukin-1. Arthritis Rheum. 52, 975–983. Data were compared by two-tailed Student t test, Wilcoxon Rank test, and by Korn, T., Bettelli, E., Gao, W., Awasthi, A., Jager, A., Strom, T.B., Oukka, M., one-way ANOVA. Where significant differences were found, the Tukey-Kramer and Kuchroo, V.K. (2007). IL-21 initiates an alternative pathway to induce multiple comparisons test was used for identifying differences between indi- proinflammatory T(H)17 cells. Nature 448, 484–487. vidual groups. Langrish, C.L., Chen, Y., Blumenschein, W.M., Mattson, J., Basham, B., Sedgwick, J.D., McClanahan, T., Kastelein, R.A., and Cua, D.J. (2005). IL-23 SUPPLEMENTAL DATA drives a pathogenic T cell population that induces autoimmune inflammation. J. Exp. Med. 201, 233–240. Supplemental Data include six figures and can be found with this article online Lochner, M., Peduto, L., Cherrier, M., Sawa, S., Langa, F., Varona, R., at http://www.cell.com/immunity/supplemental/S1074-7613(09)00327-6. Riethmacher, D., Si-Tahar, M., Di Santo, J.P., and Eberl, G. (2008). In vivo equi- librium of proinflammatory IL-17+ and regulatory IL-10+ Foxp3+ RORgamma ACKNOWLEDGMENTS t+ T cells. J. Exp. Med. 205, 1381–1393. Lockhart, E., Green, A.M., and Flynn, J.L. (2006). IL-17 production is domi- We thank J. Fletcher for discussions and B. Moran for technical help with nated by gammadelta T cells rather than CD4 T cells during Mycobacterium FACS. This work was supported by a Science Foundation Ireland PI Award tuberculosis infection. J. Immunol. 177, 4662–4669. to K.M. K.M. is cofounder and shareholder of Opsona Therapeutics Ltd, Luger, D., Silver, P.B., Tang, J., Cua, D., Chen, Z., Iwakura, Y., Bowman, E.P., a university start-up company involved in the development of anti-inflamma- Sgambellone, N.M., Chan, C.C., and Caspi, R.R. (2008). Either a Th17 or a Th1 tory therapeutics. effector response can drive autoimmunity: Conditions of disease induction affect dominant effector category. J. Exp. Med. 205, 799–810. Received: February 6, 2009 Manel, N., Unutmaz, D., and Littman, D.R. (2008). The differentiation of human Revised: May 18, 2009 T(H)-17 cells requires transforming growth factor-beta and induction of the Accepted: June 23, 2009 nuclear receptor RORgammat. Nat. Immunol. 9, 641–649. Published online: August 13, 2009 Mangan, P.R., Harrington, L.E., O’Quinn, D.B., Helms, W.S., Bullard, D.C., REFERENCES Elson, C.O., Hatton, R.D., Wahl, S.M., Schoeb, T.R., and Weaver, C.T. (2006). Transforming growth factor-beta induces development of the T(H)17 Aggarwal, S., Ghilardi, N., Xie, M.H., de Sauvage, F.J., and Gurney, A.L. (2003). lineage. Nature 441, 231–234. Interleukin-23 promotes a distinct CD4 T cell activation state characterized by McGeachy, M.J., and Cua, D.J. (2008). Th17 cell differentiation: The long and the production of interleukin-17. J. Biol. Chem. 278, 1910–1914. winding road. Immunity 28, 445–453.

340 Immunity 31, 331–341, August 21, 2009 ª2009 Elsevier Inc. Immunity gd T Cells Are an Important Source of Innate IL-17

Nagata, T., McKinley, L., Peschon, J.J., Alcorn, J.F., Aujla, S.J., and Kolls, J.K. experimental autoimmune encephalomyelitis in mice with a disrupted TCR (2008). Requirement of IL-17RA in Con A induced hepatitis and negative regu- delta chain gene. Eur. J. Immunol. 29, 4060–4071. lation of IL-17 production in mouse T cells. J. Immunol. 181, 7473–7479. Sutton, C., Brereton, C., Keogh, B., Mills, K.H., and Lavelle, E.C. (2006). Nanno, M., Kanari, Y., Naito, T., Inoue, N., Hisamatsu, T., Chinen, H., A crucial role for interleukin (IL)-1 in the induction of IL-17-producing T cells Sugimoto, K., Shimomura, Y., Yamagishi, H., Shiohara, T., et al. (2008). Exac- that mediate autoimmune encephalomyelitis. J. Exp. Med. 203, 1685–1691. erbating role of gammadelta T cells in chronic colitis of T-cell receptor alpha Takatori, H., Kanno, Y., Watford, W.T., Tato, C.M., Weiss, G., Ivanov, I.I., mutant mice. Gastroenterology 134, 481–490. Littman, D.R., and O’Shea, J.J. (2009). Lymphoid tissue inducer-like cells Nurieva, R., Yang, X.O., Martinez, G., Zhang, Y., Panopoulos, A.D., Ma, L., are an innate source of IL-17 and IL-22. J. Exp. Med. 206, 35–41. Schluns, K., Tian, Q., Watowich, S.S., Jetten, A.M., and Dong, C. (2007). Thakker, P., Leach, M.W., Kuang, W., Benoit, S.E., Leonard, J.P., and Marusic, Essential autocrine regulation by IL-21 in the generation of inflammatory S. (2007). IL-23 is critical in the induction but not in the effector phase of exper- T cells. Nature 448, 480–483. imental autoimmune encephalomyelitis. J. Immunol. 178, 2589–2598. Parham, C., Chirica, M., Timans, J., Vaisberg, E., Travis, M., Cheung, J., van Beelen, A.J., Zelinkova, Z., Taanman-Kueter, E.W., Muller, F.J., Hommes, Pflanz, S., Zhang, R., Singh, K.P., Vega, F., et al. (2002). A receptor for the het- D.W., Zaat, S.A., Kapsenberg, M.L., and de Jong, E.C. (2007). Stimulation of erodimeric cytokine IL-23 is composed of IL-12Rbeta1 and a novel cytokine the intracellular bacterial sensor NOD2 programs dendritic cells to promote receptor subunit, IL-23R. J. Immunol. 168, 5699–5708. Interleukin-17 production in human memory T cells. Immunity 27, 660–669. Park, H., Li, Z., Yang, X.O., Chang, S.H., Nurieva, R., Wang, Y.H., Wang, Y., Veldhoen, M., Hirota, K., Christensen, J., O’Garra, A., and Stockinger, B. Hood, L., Zhu, Z., Tian, Q., and Dong, C. (2005). A distinct lineage of CD4 (2009). Natural agonists for aryl hydrocarbon receptor in culture medium are T cells regulates tissue inflammation by producing interleukin 17. Nat. Immu- essential for optimal differentiation of Th17 T cells. J. Exp. Med. 206, 43–49. nol. 6, 1133–1141. Veldhoen, M., Hocking, R.J., Atkins, C.J., Locksley, R.M., and Stockinger, B. Rachitskaya, A.V., Hansen, A.M., Horai, R., Li, Z., Villasmil, R., Luger, D., (2006). TGFbeta in the context of an inflammatory cytokine milieu supports de Nussenblatt, R.B., and Caspi, R.R. (2008). Cutting edge: NKT cells constitu- novo differentiation of IL-17-producing T cells. Immunity 24, 179–189. tively express IL-23 receptor and ROR{gamma}t and rapidly produce IL-17 Volpe, E., Servant, N., Zollinger, R., Bogiatzi, S.I., Hupe, P., Barillot, E., and upon receptor ligation in an IL-6-independent fashion. J. Immunol. 180, Soumelis, V. (2008). A critical function for transforming growth factor-beta, 5167–5171. interleukin 23 and proinflammatory cytokines in driving and modulating human Roark, C.L., French, J.D., Taylor, M.A., Bendele, A.M., Born, W.K., and T(H)-17 responses. Nat. Immunol. 9, 650–657. O’Brien, R.L. (2007). Exacerbation of collagen-induced arthritis by oligoclonal, Wilson, N.J., Boniface, K., Chan, J.R., McKenzie, B.S., Blumenschein, W.M., IL-17-producing gamma delta T cells. J. Immunol. 179, 5576–5583. Mattson, J.D., Basham, B., Smith, K., Chen, T., Morel, F., et al. (2007). Devel- Shibata, K., Yamada, H., Hara, H., Kishihara, K., and Yoshikai, Y. (2007). Resi- opment, cytokine profile and function of human interleukin 17-producing dent Vdelta1+ gammadelta T cells control early infiltration of neutrophils after helper T cells. Nat. Immunol. 8, 950–957. Escherichia coli infection via IL-17 production. J. Immunol. 178, 4466–4472. Wucherpfennig, K.W., Newcombe, J., Li, H., Keddy, C., Cuzner, M.L., and Shimonkevitz, R., Colburn, C., Burnham, J.A., Murray, R.S., and Kotzin, B.L. Hafler, D.A. (1992). Gamma delta T-cell receptor repertoire in acute multiple (1993). Clonal expansions of activated gamma/delta T cells in recent-onset sclerosis lesions. Proc. Natl. Acad. Sci. USA 89, 4588–4592. multiple sclerosis. Proc. Natl. Acad. Sci. USA 90, 923–927. Yang, L., Anderson, D.E., Baecher-Allan, C., Hastings, W.D., Bettelli, E., Smith, E., Stark, M.A., Zarbock, A., Burcin, T.L., Bruce, A.C., Vaswani, D., Oukka, M., Kuchroo, V.K., and Hafler, D.A. (2008). IL-21 and TGF-beta are Foley, P., and Ley, K. (2008). IL-17A inhibits the expansion of IL-17A- required for differentiation of human T(H)17 cells. Nature 454, 350–352. producing T cells in mice through ‘‘short-loop’’ inhibition via IL-17 receptor. Zhou, L., Ivanov, I.I., Spolski, R., Min, R., Shenderov, K., Egawa, T., Levy, D.E., J. Immunol. 181, 1357–1364. Leonard, W.J., and Littman, D.R. (2007). IL-6 programs T(H)-17 cell differenti- Spahn, T.W., Issazadah, S., Salvin, A.J., and Weiner, H.L. (1999). Decreased ation by promoting sequential engagement of the IL-21 and IL-23 pathways. severity of myelin oligodendrocyte glycoprotein peptide 33 - 35-induced Nat. Immunol. 8, 967–974.

Immunity 31, 331–341, August 21, 2009 ª2009 Elsevier Inc. 341