Th17 cells are critical for the clearance of extracellular bacteria and fungi, but also contribute to the pathology of autoimmune diseases and allergic inflammation. After exposure to an appropriate cytokine environment, Th17 cells can acquire a Th1-like phenotype, but less is known about their ability to adopt Th2 and Th9 effector programs. To explore this in more detail, we used an IL-17F lineage tracer mouse strain that allows tracking of cells that formerly expressed IL-17F. In vitro–derived Th17 cells adopted signature cytokine and transcription factor expression when cultured under Th1-, Th2-, or Th9-polarizing conditions. In contrast, using two models of allergic airway disease, Th17 cells from the lungs of diseased mice did not adopt Th1, Th2, or Th9 effector programs, but remained stable IL-17 secretors. Although in vitro–derived Th17 cells expressed IL-4Rα, those induced in vivo during allergic airway disease did not, possibly rendering them unresponsive to IL-4–induced signals. However, in vitro–derived, Ag-specific Th17 cells transferred in vivo to OVA and aluminum hydroxide–sensitized mice also maintained IL-17 secretion and did not produce alternative cytokines upon subsequent OVA challenge. Thus, although Th17 cells can adopt new phenotypes in response to some inflammatory environments, our data suggest that in allergic inflammation, Th17 cells are comparatively stable and retain the potential to produce IL-17. This might reflect a cytokine environment that promotes Th17 stability, and allow a broader immune response at tissue barriers that are susceptible to allergic inflammation.
This article is featured in In This Issue, p.2613
Upon activation, naive CD4+ T cells differentiate into specific Th lineages depending on the cytokines in the environment. IL-12 promotes the IFN-γ–secreting Th1 phenotype, IL-4 induces the development of Th2 cells, which produce IL-4, IL-5, and IL-13, and the combination of IL-4 and TGF-β promotes the development of IL-9–secreting Th9 cells (1–9). Together, IL-6, TGF-β, IL-23, and IL-1β induce the development of IL-17–secreting Th17 cells (10–15). In addition to IL-17A and IL-17F, Th17 cells produce IL-21 and IL-22, and are important for immunity against extracellular bacteria and fungi, but also contribute to the pathology of autoimmune diseases and allergic inflammation (16–20). The Th17 effector program is induced by a network of transcription factors, which includes retinoic acid–related orphan receptor γt (RORγt) and STAT3, and is negatively regulated by the Th1 and Th2/Th9-inducing cytokines, IFN-γ and IL-4, respectively (11, 21–25).
Th lineages were originally thought to have stable phenotypes, and once a Th cell acquired the potential for secreting a particular cytokine, the cell was committed to this phenotype. However, experiments with Th17 cells demonstrated that they had dramatic instability, defaulting to an IFN-γ–secreting phenotype in vitro (25–28). Maintaining the Th17 phenotype in vitro requires a specific cytokine environment that includes IL-23 and IL-1 (26). The ability of a Th17 cell to acquire IFN-γ–secreting potential requires IL-12–induced STAT4 and the induction of T-bet to repress Runt-related transcription factor 1 and IFN regulatory factor 4 (IRF4) (25, 27, 29, 30). Th17 plasticity, the ability to acquire other Th cell phenotypes, is reflected by the increased expression of a stem cell signature and bivalent chromatin marks at Th lineage transcription factors that allow responsiveness to the cytokine environment (31–34). Although other Th subsets have some plasticity, the dramatic instability of the Th17 phenotype suggests that maintenance of IL-17–secreting cells might be detrimental to the host.
The plasticity of the Th17 lineage in vivo was first shown in a series of studies where polyclonal populations, or Th17 cells purified on the basis of reporter expression, were adoptively transferred into mice with autoimmune diseases including colitis and type I diabetes, or lymphopenic hosts (27, 35–37). These studies agreed with in vitro studies, and demonstrated the acquisition of IFN-γ–secreting potential after transfer. However, these studies did not exclude the possibility that some IL-17− cells could have been transferred and expanded in vivo.
The use of IL-17A and IL-17F lineage tracer mouse models allowed tracking of cells that formerly expressed IL-17, and confirmed the acquisition of a Th1-like phenotype by Th17 cells in vitro and in vivo during the development of autoimmune disease (38, 39). In experimental autoimmune encephalomyelitis (EAE), the majority of IFN-γ–secreting cells found in the CNS are former secretors of IL-17A and IL-17F (38, 39). IL-17–secreting T cells can acquire other phenotypes as well. Th17 cells adopt a follicular Th cell phenotype in Peyer’s patches inducing the development of IgA-producing germinal center B cells and promoting gut homeostasis (40). In addition, IL-17–secreting T cells can terminate IL-17 production without producing cytokines associated with other lineages. Upon clearance of acute cutaneous infection with Candida albicans, Th17 cells shut off IL-17 production, potentially dampening the immune response (39). Thus, specific inflammatory environments can induce the conversion of Th17 cells to other phenotypes in vivo, promoting homeostasis, host defense, and inflammation.
In vitro addition of IL-4 to Th17 cultures results in diminished IL-17 production and increased production of cytokines associated with the Th2 phenotype (26, 37). This suggests that Th17 cells might also acquire a proallergic phenotype in appropriate cytokine environments. However, the stability of Th17 cells in a Th2 or Th9-biased environment is not well understood. In this report, we used a newly generated IL-17F lineage tracer mouse to define the stability of IL-17–secreting T cells during allergic inflammation in vivo.
Materials and Methods
A bacterial artificial chromosome spanning the Il17f locus was used to generate a targeting vector that replaced the 3′ end of exon 1 with an enhanced GFP (EGFP)-Cre fusion protein and an frt-flanked neomycin resistance cassette (Vega Biolabs). The targeting vector was transfected into C57BL/6 embryonic stem cells. Neomycin-selected clones were screened for correct recombination, and embryonic stem cells were injected into albino C57BL/6 blastocysts. The resulting founder mice were bred to establish germline transmission. These mice are referred to as Il17f Cost mice, using Cost as an acronym for Cre On Seventeen Transcript. To excise the frt-flanked neomycin-resistance cassette, we crossed Il17f Cost mice to FLPeR mice (expressing a variant of the FLP1 recombinase gene inserted into the Gt(ROSA)26Sor locus). Il17f CostRsYFP mice were generated by crossing Il17f Cost mice to R26-stop-EYFP reporter mice (have a loxP-flanked stop sequence followed by the eYFP gene inserted into the Gt(ROSA)26Sor locus). Il17f CostRsYFP–OT-II mice were generated by crossing Il17f CostRsYFP mice to OT-II mice (express a transgenic TCR specific for chicken OVA 323–339). C57BL/6 mice were purchased from Harlan Laboratories and FLPeR mice, R26-stop-EFYP mice, and OT-II mice were purchased from Jackson Laboratory. Mice were kept in pathogen-free conditions, and all studies were approved by the Indiana University School of Medicine Animal Care and Use Committee.
Murine Th cell differentiation
Naive CD4+CD62L+ T cells were purified from spleens via magnetic isolation (Miltenyi Biotec) and activated with plate-bound anti-CD3 (2–4 μg/ml, 2C11) and soluble anti-CD28 (1–2 μg/ml). Cells were polarized to generate Th1 (5 ng/ml IL-12; 50 u/ml IL-2; 10 μg/ml anti–IL-4, 11B11), Th2 (10 ng/ml IL-4; 10 μg/ml anti–IFN-γ, XMG), Th9 (20 ng/ml IL-4; 2 ng/ml TGF-β; 10 μg/ml anti–IFN-γ), and Th17 (100 ng/ml IL-6; 2 ng/ml TGF-β; 10 ng/ml IL-23; 10 ng/ml IL-1β; 10 μg/ml anti–IL-4; 10 μg/ml anti–IFN-γ) cells. Cells were expanded after 3 d with fresh media alone for Th1 and Th2 cells or in the presence of additional cytokines for Th9 (20 ng/ml IL-4; 2 ng/ml TGF-β) and Th17 (50 ng/ml IL-6; 5 ng/ml IL-23; 5 ng/ml IL-1β) cells. Cells were harvested after 5 d in culture for analysis. For long-term Th17 cultures, cells were cultured as noted earlier for 5 d. Cells were then harvested and reactivated under long-term Th17-polarizing conditions (1 μg/ml anti-CD3; 10 ng/ml IL-23; 10 ng/ml IL-1β; 10 μg/ml anti–IL-4; 40 μg/ml anti–IFN-γ). Cells were expanded after 3 d in the presence of additional cytokines (5 ng/ml each IL-23 and IL-1β). Cells were harvested on the fifth day of the second round of culture, and live yellow fluorescent protein–positive (YFP+) cells were sorted by flow cytometry. Sorted cells were reactivated with plate-bound anti-CD3 (1 μg/ml) and cultured for a third round under Th1-, Th2-, Th9-, or long-term Th17-polarizing conditions. Cells were expanded after 3 d as noted earlier for Th1, Th2, Th9, and long-term Th17 conditions. Cells were harvested on the fifth day of the third round of culture for further analysis. For experiments using cells from Il17f CostRsYFP–OT-II mice, purified naive CD4+CD62L+ T cells were activated with soluble OVA323–339 peptide (5 μg/ml; GenScript) and soluble anti-CD28 (1 μg/ml) in the presence of CD4+ depleted splenocytes (1:5) that were first treated with mitomycin C (Calbiochem) according to the manufacturer’s instructions. Cytokines and Abs were purchased from Bio X Cell (anti-CD3, anti–IFN-γ, and anti–IL-4), BD Biosciences (anti-CD28), eBioscience (IL-1β), Miltenyi Biotec (TGF-β), Peprotech (IL-2, IL-4, IL-6, IL-12), or R&D Systems (IL-23).
For cytokine analysis from in vitro–differentiated Th cells, cells were restimulated with PMA and ionomycin for 5 h with the addition of monensin during the last 3 h of stimulation. Cells were collected, stained with a fixable viability dye, fixed with 2% final formaldehyde, permeabilized with saponin, and stained with fluorochrome-conjugated anti-mouse IFN-γ, IL-9, IL-13, IL-17A, or IL-17F. For transcription factor analysis from in vitro–differentiated Th cells, live YFP+ cells were sorted by flow cytometry. Unstimulated cells were fixed and permeabilized using a transcription factor staining buffer set (eBioscience) and stained with fluorochrome-conjugated anti-mouse GATA3, RORγt, or T-bet. For ex vivo analyses, splenocytes, bronchoalveolar lavage (BAL), lung, or mononuclear cells from the brain were stimulated with PMA and ionomycin as described earlier. Cells were collected, incubated with Fc block (BD Biosciences), and stained with fluorochrome-conjugated anti-mouse CD4. Cells were then washed, stained with a fixable viability dye and fixed, permeabilized, and stained with anti-mouse cytokine Abs as described earlier. Abs were purchased from BD Biosciences, Biolegend, or eBioscience.
Quantitative PCR was performed with total or sorted YFP+ unstimulated cells using TaqMan assays as previously described (41).
Induction of allergic airway disease
Induction of allergic airway disease (AAD) using OVA and aluminum hydroxide (alum) has been described previously (42). In brief, 8- to 10-wk-old mice were sensitized by i.p. injection of OVA (20 μg; Sigma) adsorbed with alum (2 mg; Sigma) on days 0 and 7. Mice were challenged with OVA (100 μg) intranasally (i.n.) from days 14 to 19 and were sacrificed 24 h after the last challenge. For some experiments, mice were sensitized on days 0 and 7 as mentioned earlier, and in vitro–derived OVA-specific YFP+ Th17 cells (1 × 105) were transferred by i.v. injection to the tail on day 20. Mice were challenged with OVA (100 μg) i.n. from days 21 to 26 and were sacrificed 24 h after the last challenge. Eight- to 10-wk-old mice were administered intranasal (i.n.) doses of HDM (50 μg; Greer) for 3 consecutive days each week for 5 wk, to induce AAD using house dust mite (HDM). Mice were sacrificed 24 h after the last dose was administered. For both models, mice were sacrificed by i.p. injections of pentobarbital (5 mg/mouse) 24 h after the final i.n. challenge. The trachea was cannulated and the lungs were lavaged three times with 1 ml PBS. The cells recovered in the BAL fluid were counted with a hemacytometer. The lungs were isolated and single-cell suspensions were prepared using a gentleMACS Dissociator (Miltenyi Biotec).
Induction of EAE
Induction of EAE disease has been previously described (43, 44). In brief, 8- to 10-wk-old female mice were immunized s.c. on days 0 and 7 with myelin oligodendrocyte glycoprotein (MOGp35–55) Ag peptide (100 μg; Genemed Synthesis) emulsified in CFA (150 μl; Sigma). Mice were injected i.p. with pertussis toxin (100 ng; Sigma) on days 0 and 2. Mice were sacrificed 19 d after induction of disease, and spleen and brain were harvested. Mononuclear cells were isolated from brain using a 30%/70% Percoll gradient. Splenocytes and mononuclear cells from the brain were stimulated for 5 h with PMA and ionomycin with the addition of monensin during the final 3 h of stimulation before further staining for flow cytometry analysis.
The one-way ANOVA was used for statistical comparison. The p values ≤0.05 were considered significant.
Generation and characterization of Il17f Cost mice
IL-17 reporter and lineage tracer mouse strains are tools to identify cells that currently produce IL-17 (reporter mouse) or that have produced IL-17 (lineage tracer mouse). Our goal was to generate a mouse strain that combined both approaches, to simultaneously identify cells that currently produce IL-17F or that previously produced IL-17F, but have since turned off the Il17f gene. An EGFP-Cre recombinase fusion was inserted into exon 1 of Il17f, termed Cost (Fig. 1A). Il17f Cost mice express a fusion of EGFP and Cre recombinase when the Il17f locus is turned on. To observe Cre activity, we crossed Il17f Cost mice with a reporter mouse strain that expresses YFP upon expression of Cre (referred to as RsYFP in this article). In theory, cells from Il17f CostRsYFP mice would express EGFP and Cre when Il17f is turned on, and retain Cre-induced YFP expression if Il17f expression were discontinued. Expression of EGFP and YFP was tested from in vitro–derived Th17 cells from Il17f CostRsYFP mice, and we found that although both EGFP and YFP could be detected in Th17 cells by immunofluorescence microscopy, YFP, but not EGFP, was detected by flow cytometry, possibly because of insufficient expression of the fusion protein for cytometric detection (data not shown). Thus, in subsequent experiments, Il17f CostRsYFP mice were used as an Il17f lineage tracer mouse strain, similar to that previously described by Croxford et al. (36), using flow cytometry analyses to define YFP+ cells as those that had activated the Il17f locus.
To validate the Il17f CostRsYFP mouse strain, we isolated naive CD4+ T cells from the spleen of Il17f +/+RsYFP/YFP, Il17f +/CostRsYFP/YFP, and Il17f Cost/CostRsYFP/YFP mice and stimulated them under Th17-polarizing conditions for 5 d. As expected, Th17 cells from Il17f +/+RsYFP/YFP mice do not express YFP, but express high percentages of IL-17F+ and IL-17A+ cells. Th17 cells from Il17f +/CostRsYFP/YFP mice demonstrate expression of YFP and display an expected reduction in IL-17F, because the cells have only one functioning Il17f allele. Th17 cells derived from Il17f Cost/CostRsYFP/YFP mice display a further increase in YFP expression but do not produce IL-17F, because they are Il17f null (Fig. 1B). Although only the Il17f locus was targeted in the generation of Il17f Cost mice, it is important to note that Th17 cells derived from Il17f +/CostRsYFP/YFP and Il17f Cost/CostRsYFP/YFP mice produce less IL-17A than control mice, suggesting that there is an effect of the Cost allele on the expression of the adjacent Il17 allele (Fig. 1B). Importantly, in comparison with total Th17 cells derived from Il17f +/CostRsYFP/YFP and Il17f Cost/CostRsYFP/YFP mice, the YFP+ population of Th17 cells has a higher expression of IL-17A and IL-17F, demonstrating an enrichment of IL-17–producing cells in the YFP+ population (Fig. 1B).
To confirm that in vitro–differentiated Th17 cells and not other Th subsets expressed YFP, we isolated naive CD4+ T cells from the spleen of Il17f +/CostRsYFP/YFP mice and stimulated them under Th1, Th2, Th9, or Th17-polarizing conditions. YFP is expressed in Th17 cells, but not in other Th lineages (Fig. 1C). To further validate the use of Il17f CostRsYFP mice for in vivo studies intended to track the fate of IL-17F–expressing cells, we induced EAE in Il17f +/CostRsYFP/YFP mice and analyzed cytokine secretion in CD4+ cells from the brain. Similar to what has been previously reported (38, 39), we found that CD4+ cells from the brain of mice at the peak of paralysis coexpressed IL-17A and IFN-γ. The cytokine-positive population was enriched in the CD4+YFP+ population, demonstrating that in EAE, some IFN-γ–secreting CD4+ cells coexpress IL-17F or arise from an IL-17F+ precursor (Fig. 1D). Together, these data demonstrate that Il17f CostRsYFP mice can be used for in vitro and in vivo studies to further explore the stability of IL-17F–secreting cells.
Altered Th17 cytokine expression upon stimulation under Th1-, Th2-, or Th9-polarizing conditions
Th17 cells can produce robust amounts of IFN-γ when stimulated in vitro under Th1-skewing conditions; however, less is known about their cytokine profile when stimulated under Th2- or Th9-polarizing conditions. To examine this, we isolated naive CD4+ T cells from the spleen of Il17f +/CostRsYFP/YFP mice and stimulated them under Th17-polarizing conditions for two rounds of culture to establish a more differentiated population of Th17 cells. YFP+ cells were sorted and maintained under Th17-polarizing conditions or stimulated under Th1-, Th2-, or Th9-polarizing conditions for a third round of culture. After two rounds of stimulation under Th17-polarizing conditions, YFP+ cells produce IL-17A and IL-17F with little to no IFN-γ, IL-13, or IL-9 production (Fig. 2A). In most experiments, we observed diminished IL-17F expression after the second round of polarization. YFP+ cells stimulated for a third round of culture under Th17-polarizing conditions display an increase in IL-17A+ and IL-17A+IL-17F+ cells (Fig. 2B, 2C). In contrast with cells maintained under Th17-skewing conditions for a third round of culture, YFP+ cells stimulated under Th1-, Th2-, or Th9-polarizing conditions display reduced IL-17A production, particularly in the IL-17A+IL-17F+ population (Fig. 2B, 2C). YFP+ Th17 cells stimulated under Th1-polarizing conditions display reduced IL-17A+IFN-γ− cells with an increase in IL-17A+IFN-γ+ and IL-17A−IFN-γ+ cells in comparison with cells maintained under Th17 conditions (Fig. 2B, 2D). An increase in IL-17A+IL-13+ and IL-17A−IL-13+ cells is observed when YFP+ Th17 cells are cultured under Th2-skewing conditions, and to a lesser extent after culture under Th9-polarizing conditions (Fig. 2B, 2E). Furthermore, YFP+ Th17 cells cultured under Th9-polarizing conditions display an increase in IL-17A+IL-9+ cells along with a smaller increase in IL-17A−IL-9+ cells compared with cells maintained under Th17-skewing conditions (Fig. 2B, 2F). YFP+ Th17 cells cultured under Th2-polarizing conditions also display an increase in IL-17A−IL-9+ cells with a more moderate increase in IL-17A+IL-9+ cells; however, both populations are smaller than in cells cultured under Th9-polarizing conditions (Fig. 2B, 2F). Enhanced IL-9 secretion from Th17 cells stimulated under Th2-skewing conditions could be an effect of prior TGF-β–induced signals during Th17 differentiation in combination with additional IL-4 stimulation under Th2-polarizing conditions. Taken together, these data demonstrate that Th17-associated cytokines are differentially expressed when YFP+ Th17 cells are stimulated under different Th-polarizing conditions, whereas the expression of IFN-γ, IL-13, or IL-9 is increased when YFP+ Th17 cells are stimulated under Th1-, Th2-, or Th9-polarizing conditions, respectively.
Stimulation of Th17 cells under different Th-polarizing conditions induces the expression of the respective Th-associated transcription factors
The changes in cytokine production observed in YFP+ Th17 cells maintained under Th17-polarizing conditions or stimulated under Th1-, Th2-, or Th9-polarizing conditions led us to determine whether there were also alterations in transcription factor expression among these cell populations. To examine this, we cultured naive CD4+ T cells from the spleen of Il17f +/CostRsYFP/YFP mice as described for Fig. 2 and sorted YFP+ cells after 3 d of the third round of stimulation under Th17-skewing conditions or Th1-, Th2-, or Th9-polarizing conditions. Although we saw a reduction in IL-17A+ cells when YFP+ Th17 cells were cultured under Th1-, Th2-, or Th9-skewing conditions in comparison with cells maintained under Th17-polarizing conditions (Fig. 2), RORγt protein was similar among YFP+ Th17 cells cultured under different polarizing conditions (Fig. 3A, 3B). However, in comparison with YFP+ Th17 cells maintained under Th17-skewing conditions, Rorc mRNA was significantly reduced in cells stimulated under Th1-polarizing conditions (Fig. 3C).
YFP+ Th17 cells stimulated under Th1-polarizing conditions express increased T-bet protein and mRNA (Fig. 3), coinciding with the observed increase in IFN-γ production from these cells (Fig. 2). GATA3 is essential for Th2 and Th9 development, and YFP+ Th17 cells stimulated under Th2- or Th9-polarizing conditions display enhanced GATA3 levels in comparison with cells maintained under Th17-polarizing conditions (Fig. 3). Furthermore, Th17 cells stimulated under Th9-polarizing conditions express a significant increase in Irf4 and Maf, both of which encode transcription factors with enriched expression in Th9 cells (Fig. 3C). BATF has been shown to play an important role in the development of a number of Th subsets, and although Batf is expressed at similar levels in YFP+ Th17 cells stimulated under Th2-, Th9-, or Th17-polarizing conditions, expression levels are significantly reduced in YFP+ Th17 cells stimulated under Th1-polarizing conditions in comparison with YFP+ Th17 cells maintained under Th17-skewing conditions (Fig. 3C). These data demonstrate that YFP+ Th17 cells stimulated under Th1-, Th2-, or Th9-polarizing conditions induce the expression of transcription factors associated with the development of each Th subset.
Th17 cell stability in acute AAD
Several reports have demonstrated that Th17 cells can adopt the IFN-γ–secreting phenotype of Th1 cells in vivo, and we have confirmed that using our reporter mouse (Fig. 1D). However, it is less clear whether Th17 cells can adopt a Th2 or Th9 effector program in an in vivo environment that promotes their development. To explore this possibility, we first examined the stability of Th17 cells in an acute model of AAD. Il17 f+/CostRsYFP/YFP mice were sensitized with OVA and alum followed by i.n. challenges with OVA. In comparison with control mice challenged with PBS, mice challenged with OVA display an increase in the total number of cells in the BAL, as well as in the number of CD4+YFP+ cells isolated from the BAL (Fig. 4A). Although most of the CD4+ T cells from the BAL are single producers of IL-17A, IFN-γ, or IL-13, there is a small population of cells that coproduce IL-17A and IFN-γ or IL-17A and IL-13 (Fig. 4B). CD4+YFP+ cells produce IL-17 and generate a proliferative response to OVA (data not shown), demonstrating that YFP+ cells are Ag specific. However, the IL-17A–producing CD4+YFP+ cells from the BAL and lung do not express the other Th cell–associated cytokines analyzed (Fig. 4B, 4C, data not shown), suggesting a stable Th17 phenotype in acute AAD.
Th17 cell stability in chronic AAD
Because there are differences in the allergic environment induced by different models of AAD, we also explored the stability of Th17 cells in a chronic model of AAD using HDM challenge, a model that has been shown to be dependent on IL-17 and Th17 cells (45–47). Il17f +/CostRsYFP/YFP mice received three consecutive i.n. challenges with HDM each week for 5 wk. Allergic mice display an increase in the total number of cells that infiltrated the BAL along with an increase in CD4+YFP+ cells in the BAL in comparison with nonallergic control mice (Fig. 5A). Similar to what was observed in the acute model of AAD, CD4+YFP+ cells from the BAL and lung of allergic mice express IL-17A, but do not express IFN-γ, IL-13, or IL-9 (Fig. 5B, data not shown). These data demonstrate that YFP+ Th17 cells remain IL-17 producers and do not express cytokines associated with other Th subsets during the development of acute and chronic AAD.
Differential cytokine receptor expression from in vitro– and in vivo–derived Th17 cells
Our data show that Th17 cells derived in vitro are capable of adopting a Th2 or Th9 effector program; however, in the models of AAD tested, Th17 cells remain IL-17 secretors and do not adopt alternative Th phenotypes. It is possible that in vitro– and in vivo–derived Th17 cells display differences in cytokine receptor expression, affecting their responsiveness to cytokines in the environment. To explore this further, we first analyzed cytokine receptor expression from sorted YFP+ Th17 cells derived from Il17f+/CostRsYFP/YFP mice before stimulation under Th-polarizing conditions (Th17 preswitch). As expected, after two rounds of culture under Th17-polarizing conditions, YFP+ cells display significantly increased levels of Il23r expression in comparison with naive CD4+ T cells (Fig. 6A). YFP+ Th17 cells display higher levels of Il12rb2 and Il4ra expression compared with littermate control naive CD4+ T cells (Fig. 6A). YFP+ Th17 cells before switching to Th2 conditions have high IL-4Rα staining on a similar percentage of cells to cells in Th2 cultures (Fig. 6B). We next assessed cytokine receptor expression from CD4+ T cells in the BAL and lung of Il17f +/CostRsYFP/YFP mice that had developed OVA and alum-induced AAD. CD4+YFP− T cells from the BAL and lung of allergic mice display enhanced expression of IL-4Rα in comparison with control mice, which coincides with an increase in Th2 and Th9 cell development in allergic mice (Figs. 6C and 4B, data not shown). However, CD4+YFP+ T cells from the BAL and lung of allergic mice do not express IL-4Rα (Fig. 6C, data not shown). Together, these data suggest that although in vitro–derived Th17 cells express the cytokine receptors necessary to respond to cytokines essential for the development of other Th effector programs, Th17 cells developed during AAD do not.
In vitro–derived Ag-specific Th17 cells remain stable IL-17 secretors in an in vivo allergic environment
To understand whether the differences observed in the stability of Th17 cells derived in vitro or during the development of AAD is due to differences in cytokine receptor expression or can be attributed to the cytokine environment induced during AAD, we determined whether in vitro–derived, Ag-specific Th17 cells could adopt other Th effector programs upon transfer to an allergic environment. Ag-specific naive CD4+ T cells from Il17f +/CostRsYFP/YFP–OT-II mice were stimulated under Th17-polarizing conditions for two rounds of culture. YFP+ Th17 cells were sorted, transferred i.v. to OVA and alum-sensitized wild type mice, and mice were subsequently challenged with OVA. We transferred a relatively small number of cells (105) to be able to track the cells without having the transferred cells dominate the in vivo response. The overall cellular infiltrate in the lung was not different between recipient mice that did or did not receive YFP+ Th17 cells (data not shown). Whereas endogenous CD4+YFP− cells from the BAL and lung express IL-17A, IFN-γ, or IL-13, transferred CD4+YFP+ cells maintain expression of IL-17A, but do not produce IFN-γ, IL-13, or IL-9 (Fig. 7, data not shown). These data demonstrate that whereas in vitro–derived Th17 cells express IL-4Rα, respond to IL-4–induced signals, and adopt proallergic phenotypes in vitro, Th17 cells in a Th2- and Th9-biased proinflammatory environment retain an IL-17–secreting phenotype without adopting alternative Th effector programs.
Lineage differentiation and commitment are essential biological processes, which have been studied extensively in a number of systems, including hematopoiesis and T cell development. Recent research has demonstrated that Th lineages display considerable plasticity and can change their pattern of expression of lineage-specific transcription factors and cytokines in response to an altered cytokine environment. Th17 cells are unstable in vitro, and the development of lineage tracer mouse models has further revealed an unstable phenotype in vivo, particularly in inflammatory environments that promote Th1 development. Our results demonstrate that Th17 cells are more stable in a proallergic environment, not acquiring other cytokine-secreting phenotypes and maintaining an IL-17–secreting phenotype.
The mechanism of Th17 stability in allergic inflammation is at least 2-fold. First, Th17 cells derived in vivo lack expression of the IL-4Rα-chain that would be required for the induction of either Th2 or Th9 phenotypes. This is distinct from Th17 cells cultured in vitro, and suggests that in vivo, there is an active process that represses expression of Il4ra. Second, the allergic environment in the lung was not sufficient to promote Th17 switching to Th2 or Th9 phenotypes. Even in vitro–derived Th17 cells are incapable of acquiring IL-13 or IL-9 secretion when adoptively transferred into challenged mice. This could simply be the result of an insufficient concentration of IL-4 to promote a switch. Although it is difficult to determine what concentration of cytokine a Th17 cell might encounter in vivo, it is clearly sufficient in these experiments to generate Th2 cells from the endogenous T cell population. It is also possible that additional components of the allergic inflammatory milieu actively contribute to Th17 stability. Distinguishing among these possibilities will require further studies.
Allergic inflammation has a heterogeneous cellular infiltrate and is orchestrated by a large number of immune cells, including Th1, Th2, Th9, and Th17 cells (48). Two recent reports have identified a population of T effector/memory cells in humans that coexpress GATA3 and RORγt, and coproduce Th2 and Th17 cytokines. This cell population was increased in asthmatic patients compared with nonasthmatics and further identified in the lungs of allergic mice; however, the origin of the Th2/Th17 effector/memory population is not completely understood (49, 50). In our experiments, we detected a small population of CD4+ cells in the lung of allergic mice that coproduced IL-17A and IL-13, although these cells did not express YFP. This would suggest that IL-17A/IL-13–secreting T cells did not previously express IL-17F or that they only recently expressed IL-17F, and YFP expression was not yet established. Our results imply that IL-17A/IL-13–secreting T cells might develop from cells with a Th2 phenotype but are unlikely to develop from Th17 cells.
Th17 cells play an important role in the development of allergic inflammation in the lung of mice by promoting neutrophilic airway inflammation and further enhancing Th2-mediated airway eosinophilia (51, 52). Th17 cells also promote steroid-resistant airway responses in mice, linking them to severe AAD (53). However, IL-17 may have a dual role in allergic inflammation because it is essential during allergic sensitization in some models of airway inflammation, but it has also been shown to repress airway responses in mice that have been sensitized to allergen (54–56). The role of IL-17 in human allergic disease is less clear. Although IL-17 levels are increased in the lung of asthmatic patients, the cellular source of IL-17 and its association with disease severity are still not well defined (57–60). The association of asthma and single nucleotide polymorphisms in IL1R1 and RORA, and clinical effects of trials targeting IL-17A and IL-17R in asthma, are suggestive of a role for the Th17 cell pathway in human allergic disease (20, 61–63). Collectively, these findings suggest a role for Th17 cells in the development of allergic responses in the mouse and human lung, and it remains possible that Th17 cell stability is essential for the development of allergic disease.
The stability of Th17 cells in a proallergic environment might be important for the development of appropriate immune responses. Hirota et al. (39) found that during the development of a skin infection with Candida albicans, Th17 cells repressed IL-17 expression and did not secrete alternative cytokines upon clearance of the infection, which was accompanied by diminished IL-23 expression and increased IL-10 expression in the skin, suggesting a switch to an anti-inflammatory environment (39). The role in the immune response of Th17 (YFP+) cells that have turned off IL-17 expression but do not acquire the ability to secrete other cytokines is not well defined. These cells could simply be a result of turning off a specific arm of the immune system. In contrast, these cells might perform a distinct function in vivo, or acquire distinct cytokine-secreting phenotypes at a different time point. Our studies did not address the kinetics of YFP and IL-17 expression over time in allergic mice, although the level of IL-17 produced by in vitro–derived YFP+ Th17 cells remained the same before and after transfer to allergic mice. Moreover, we observed only a small number of YFP+ cells in the draining lymph node, suggesting that switched cells were not migrating to other lymphoid organs. Our results suggest a selective advantage for Th17 stability in this environment. The proallergic cytokine milieu might maintain the Th17 phenotype to potentiate neutrophilic inflammation that might aid in immunity to infection coincident with a flare of allergic inflammation. Thus, in contrast with cytokine mileaux promoting an evolving immune response from Th17-mediated to Th1-mediated inflammation, allergic inflammation might allow for a broader immune response to potential pathogens.
Using our newly generated Il17f Cost mice, we demonstrated that although Th17 cells can adopt the effector programs of Th1, Th2, or Th9 cells in vitro, during allergic inflammatory disease, Th17 cells are comparatively stable and retain the potential to produce IL-17. Thus, our data demonstrate that the inflammatory environment dictates the stability of Th17 cells in vivo. Future studies will define inflammatory mediators and underlying mechanisms involved in the regulation of Th17 cell stability in the allergic airway environment and other anatomic locations of allergic inflammation.
This work was supported by Public Health Service Grants AI045515 (to M.H.K.), T32HL007910 (to D.P. and N.L.G.-B.), T32 AI060519 (to G.L.S.), T32CA111198 (to Q.Y.), T32HL091816 (to Q.Y.), and F31 Al100542 (to O.A.) and by the Herman B Wells Center, which was supported by the Riley Children’s Foundation.
Abbreviations used in this article:
allergic airway disease
Cre On Seventeen Transcript
experimental autoimmune encephalomyelitis
house dust mite
IFN regulatory factor
retinoic acid–related orphan receptor
yellow fluorescent protein.
The authors have no financial conflicts of interest.