Prostaglandin I₂ promotes the development of IL-17-producing γδ T cells that associate with the epithelium during allergic lung inflammation

Abstract

γδ T cells rapidly produce cytokines and represent a first line of defence against microbes and other environmental insults at mucosal tissues and are thus thought to play a local immunoregulatory role. We show that allergic airway inflammation was associated with an increase in innate IL-17-producing γδ T (γδ-17) cells that expressed the αEβ7 integrin and were closely associated with the airway epithelium. Importantly, prostaglandin (PG)I₂ and its receptor IP, which downregulated airway eosinophilic inflammation, promoted the emergence of these intraepithelial γδ-17 cells into the airways by enhancing IL-6 production by lung eosinophils and dendritic cells. Accordingly, a pronounced reduction of γδ-17 cells was observed in the thymus of naïve mice lacking the PGI₂ receptor IP, as well as in the lungs during allergic inflammation, implying a critical role for PGI₂ in the programming of “natural” γδ-17 cells. Conversely, iloprost, a stable analog of PGI₂, augmented IL-17 production by γδ T cells but significantly reduced the airway inflammation. Together, these findings suggest that PGI₂ plays a key immunoregulatory role by promoting the development of innate intraepithelial γδ-17 cells through an IL-6-dependent mechanism. By enhancing γδ-17 cell responses, stable analogs of PGI₂ may be exploited in the development of new immunotherapeutic approaches.

Introduction

In response to an infection, allergen or environmental insult, various cells of the innate and adaptive immune systems participate in a coordinated fashion to mold an appropriate mucosal response. Three distinct subsets of CD4⁺ Th cells, Th1, Th2, and Th17 cells, play critical roles during adaptive immune responses. These effector T cells not only recruit and activate CD8⁺ T cells, neutrophils, eosinophils, macrophages and other effector cells, but also act directly on several cells in the mucosal site, with the epithelial cells typically playing a key role. The diverse functions of CD4⁺ T cells are typically mediated by their cytokine secretion patterns. Th1 cells secrete IFN-γ and are essential for controlling intracellular pathogens and promoting IgG2a production, whereas Th2 cells, which produce IL-4, IL-5 and IL-13, are critical for IgE production, eosinophil recruitment and clearance of helminth infections (1). Th2 cells are also responsible for driving the inflammatory response in asthma and other allergic diseases (2). Conversely, Th17 cells, which produce a distinctive set of cytokines including IL-17 (also known as IL-17a), IL-17f, IL-21 and IL-22, play crucial roles during immune responses against diverse pathogens including extracellular bacteria and fungi (3, 4). The development of Th17 cells is critically dependent on TGF-β and IL-6. Importantly, IL-6 is effective at suppressing the TGF-β-induced generation of Foxp3⁺ regulatory T cells (5, 6). IL-17 mediates several important effects in the mucosal site by acting on epithelial cells to induce the expression of chemokines that recruit neutrophils (7), antimicrobial peptides such as β-defensins (8), and the polymeric immunoglobulin receptor (9, 10). However, γδ T cells can also produce IL-17 and these innate IL-17-producing T cells are involved in sensing stress, injury or pathogens and serve an immunoregulatory role at epithelial sites (11, 12). In the lung, the Vγ4⁺ T cell subset has been shown to be induced during allergic lung inflammation and play an important role in limiting inflammation (13, 14). In marked contrast, the effector CD4⁺ T cells drive the mucosal inflammation and promote airway remodeling if left unregulated.

Emerging evidence suggests that prostanoids, such as prostaglandin (PG)I₂, PGE₂ and PGD₂, play prominent roles in promoting and regulating immune responses by facilitating Th1 differentiation, amplifying IL-23-mediated Th17 expansion (15–17), regulating innate iNKT cell responses (18) and suppressing Th2-mediated responses (19–21). Prostanoids, a family of lipid mediators consisting of the prostaglandins and thromboxanes, are generated from arachidonic acid in response to various stimuli by the sequential actions of cyclooxygenase 1 or 2, and the respective synthases. Prostaglandins are abundant at sites of inflammation and mediate their actions mainly through their respective G-protein coupled receptors that include the PGI₂ receptor (IP), PGD₂ receptor (DP) and PGE₂ receptor subtypes (EP1-EP4) (22). Each of these prostanoids, acting through its own receptor(s), differs in its effect on cyclic AMP (23) and regulation of pathophysiological processes. PGI₂ is highly labile and rapidly hydrolyzed at physiological pH to form the inactive 6-keto-PGF1α (24). Studies using mice deficient in the PGI₂ receptor IP (IP−/− mice) provided strong evidence that PGI₂ (also known as prostacyclin) exerts multiple effects in promoting as well as suppressing immune and inflammatory responses (17, 19, 20, 25, 26). Specifically, PGI₂ is known to suppress allergic airway inflammation and remodeling in models of asthma (19, 21, 27).

Many of the inflammatory characteristics of allergic asthma, namely pulmonary eosinophilia, IgE production, mucus hypersecretion and airway hyperresponsiveness, are mediated by CD4⁺ Th2 cells. It has become increasingly evident that airway epithelial cells play a central role in the inflammatory response. Specifically, prolonged activation of epithelial cells promotes mucus production and stimulates fibrogenic and remodeling processes (28–31). However, the interaction of innate and adaptive immune cells with the airway epithelium and the impact of the inflammatory response on this collaboration are poorly understood. The αEβ7 integrin, which serves as a receptor for E-cadherin (32), is expressed by intestinal T cells and by T cells in the lung that associate with airway epithelial cells (33). TGF-β is known to play an important role in the induction of both IL-17 by T cells (4) and the expression of the αEβ7 integrin (34). In the present study, we demonstrated that an innate γδ-17 cell response develops in tandem with the allergic lung inflammation and that these cells expressed the αEβ7 integrin and are closely associated with the epithelium. Given that PGI₂ plays an important immunoregulatory role during allergic lung inflammation (20, 21), the aim was to examine the involvement of this prostanoid in molding the γδ T cell response. We found that PGI₂-IP signaling is essential in promoting the development of intraepithelial γδ-17 cells. Our observation that naïve IP−/− mice lacked γδ-17 cells in the thymus is a clear demonstration that PGI₂ plays an important role in the development of “natural” γδ-17 cells. The production of IL-17 by γδ T cells was dependent on IL-6 whose expression was promoted by PGI₂. Given the role of γδ T cells in epithelial repair and homeostasis, it is conceivable that promoting responses by γδ-17 cells using stable analogs of PGI₂, such as iloprost, may provide a novel approach in resolving some of the epithelial changes that occur during allergic airway inflammation.

Materials and Methods

Mice

All mice were maintained in microisolator cages and treated in accordance with NIH guidelines and the American Association of Laboratory Animal Care regulations. All animal experiments were approved by the University of Montana Institutional Animal Care and Use Committee. Female or male C57BL/6, BALB/c (Jackson Laboratory), IP−/− and DO11.10 transgenic mice and were used throughout (6–8-wk-old). IP−/− mice were developed and provided by Dr. Garret A. FitzGerald Laboratory, Institute for Translational Medicine and Therapeutics, Philadelphia, PA. DO11.10 transgenic mice (originally developed by Dr. D. Y. Loh, Howard Hughes Medical Institute, St. Louis, MO and provided by Dr. Ethan Shevach (National Institutes of Health, Bethesda, MD) were bred under pathogen-free conditions in a barrier facility.

Media

Cells were cultured in RPMI 1640 media supplemented with 10% fetal bovine serum, L-glutamine (Gibco, Carlsbad CA), penicillin and streptomycin (Gibco), HEPES (Gibco), Sodium Pyruvate (Gibco) and 2-mercaptoethanol (Sigma).

Preparation of DO11.10 CD4⁺ Th2 cells

To generate CD4⁺ Th2 effector cells, peripheral lymph nodes (PLN) obtained from DO11.10 mice were first depleted of CD8⁺ cells using MACS beads (Miltenyi Biotech Inc., Auburn, CA, USA) and incubated (5 × 10⁵/ml) in complete RPMI media for 4 days in the presence of OVA_323–339 peptide (1 µg/ml, Mimotopes, San Diego, CA, USA), and IL-4 (2 ng/ml, R&D Systems, Minneapolis, MN, USA) plus monoclonal antibody (mAb) anti-IFN-γ (5 µg/ml R4-6A2, American Type Tissue Collection (ATTC), Manassas, VA, USA). After 4 days of incubation, cells were restimulated using identical culture conditions as previously but this time in the presence of exogenous IL-2 (10 ng/ml, R&D Systems) for a further 4 days. On day eight, the cells were depleted of class II⁺ cells by panning by incubating with anti-class II mAb (5 µg/ml M5/114, ATCC) for 30 minutes, then plate-bound mouse anti-rat IgG (10 µg/ml, Jackson Immuno, West Grove, PA) for one hour. Non-adherent (Class II⁻ cells) CD4⁺ Th2 cells were collected for analysis.

Transfer of polarized DO11.10 CD4⁺ Th2 cells and OVA challenge

Eight-day polarized DO11.10 CD4⁺ Th2 cells (10⁷ cells/mouse that are devoid of γδ T cells) were adoptively transferred into BALB/c animals by i.v. injection. Mice (four to six per group) were then challenged by exposure in a chamber to aerosolized solutions of OVA (0.5%, Grade V, Sigma-Aldrich) for 20 min/day, over 7 consecutive days using a Wright’s nebulizer. Control mice were exposed to OVA aerosols but did not receive DO11.10 Th2 cells.

Intranasal administration of iloprost, anti-IL-6 or anti-Vγ4

In certain experiments, mice were treated intranasally with iloprost, anti-IL-6 mAb, or anti-Vγ4 mAb (four instillations during the aerosol challenge, i.e. every 48 h) to examine the effect on Th2-mediated inflammation. Briefly, DO11.10 mice were given Th2 cells (6 × 10⁶ cells/ml) and then challenged with aerosolized OVA for 7 consecutive days. On days 0, 2, 4 and 6 of OVA inhalation, Th2 recipient mice were lightly anesthetized with isofluorane to allow intranasal administration of iloprost (2.5 µg, Cayman chemical, Ann Arbor, Michigan), anti-IL-6 mAb (clone MP5-20F3, 15 µg, BioLegend), anti-Vγ4 mAb (25 µg UC3, gift from Dr Jeffrey A. Bluestone, University of California) in 30 µl PBS or sterile PBS alone (vehicle).

OVA immunization and challenge of mice

C57BL/6 WT and IP−/− mice were immunized with OVA (20 µg) using an alum adjuvant (100 µg) by intra-peritoneal injection and after 10 days the mice were exposed to either OVA aerosols or PBS (control) for 20 min/day over 7 consecutive days.

Level of pulmonary inflammation

Following OVA inhalation for 7 days (i.e. on day 8), bronchoalveolar lavage (BAL) was performed (3 × 0.5 ml of PBS) to collect BALF for analysis. EPO levels in the BAL cells were determined by colorimetric analysis. Cell differential percentages were determined by light microscopic evaluation of Hema3-stained cytospin preparations and expressed as absolute cell numbers. Lung tissue was dispersed by collagenase (Type IV, Sigma-Aldrich) and lung mononuclear cells (LMC) were isolated by Percoll (Sigma) density gradient for functional analysis.

Flow cytometry and intracellular IL-17, IL-6, IL-4 and IFN-γ staining

FACSAria II (BD Biosciences, San Jose, CA) was used to enumerate the number of αβ⁺ T cells γδ⁺ T cells, KJ1-26⁺ T cells (KJ-126 is an antibody that recognizes I-Ad restricted TCR specific for OVA), Gr-1⁺ (Gr-1 is GPI-linked myeloid differentiation marker also known as Ly-6G expressed by granulocytes and monocytes), CD11b⁺, CCR3⁺, CD11c⁺ and Class II⁺ cells in the LMC, BALF, spleen or thymus co-expressing the αEβ7 integrin (a member of the integrin family expressed almost exclusively by cells of the T lymphocyte lineage in mucosal tissues), IL-17, IL-6, IL-4, IFN-γ, CD4, CD8, Vγ4, Vγ5 or Vγ1 using specific mAb that include anti-αβ TCR (H57, FITC-conjugated, BD Biosciences), anti-γδ TCR (GL3, allophycocyanin (APC)-conjugated, BD Biosciences), anti-mouse DO11.10 TCR (KJ1-26, APC-conjugated, eBioscience, San Diego, CA), anti-αE (anti-CD103, APC-conjugated, BioLegend, San Diego, CA) and anti-integrin β7 (M293, PE-conjugated, BD Biosciences), anti-CD4 (GK1.5, APC-Cy7-conjugated, BD Biosciences), anti-CD8 (53-6.7, PE-conjugated, BD Biosciences), anti-Vγ4 (UC3, FITC-conjugated, gift from Dr. J. A. Bluestone), anti-Vγ5 (F536, FITC-conjugated, gift from Dr J. A. Bluestone), anti-Vγ1 (2.11, FITC-conjugated, BioLegend), anti-Gr-1 (Ly-6G/Ly-6C, RB6-8C5, APC-Cy7-conjugated, BioLegend), anti-CD11b (M1/70, APC-conjugated, BioLegend), anti-CCR3 (CD193, TG14/CCR3, Alexa Fluor 647-conjugated, BioLegend), anti-CD11c (N418, APC-conjugated, BioLegend), and anti-Class II (I-A/I-E, M5/114.15.2, APC-conjugated, BioLegend). Spleen and thymus cells were first enriched by depletion of CD4⁺ and CD8⁺ cells (and also B220⁺ cells in the spleen cells) using MACS beads (Miltenyi Biotech). CD11b⁺CCR3⁺ cells were sorted using FACSAria II (BD Biosciences), and cytospin preparations were stained with Hema3 to confirm eosinophil purity by light microscopic evaluation (>95%).

For analysis of intracellular IL-17, IL-6, IL-4 or IFN-γ, cells (1 × 10⁶) were stimulated with 50 ng/ml PMA plus 500 ng/ml ionomycin (Sigma) in the presence of 1 µl of the protein transport inhibitor BD GolgiPlug containing brefeldin A (BD Biosciences) for 4 h at 37 °C (this step was omitted for IL-6 staining). Cells (0.5 × 10⁶) were then blocked using 2.4G2 supernatant (ATCC) and stained with the appropriate conjugated mAb or isotype control. Following treatment with fixation and permeabilization buffers (BioLegend), cells were intracellularly stained using the following antibodies (all from BioLegend): APC or PE-conjugated anti-IL-17 (clone TC11-18H10.1), PE-conjugated anti-IL-6 (MP5-20F3), APC-conjugated anti-IL-4 (clone 11B11) or FITC-conjugated anti-IFN-γ (clone XMG1.2) and analyzed by FACSAria II.

Measurement of IL-17, IL-4 and IFN-γ production

To measure IL-17 production, LMC or spleen cells were stimulated with anti-CD3 (2 µg/ml 2C11, from ATCC), anti-mouse TCR β chain (5 µg/ml H57-597), anti-mouse TCR Vγ4 (5 µg/ml UC3), anti-mouse TCR Vγ5 (5 µg/ml, F536), anti-mouse γδ TCR (5 µg/ml UC7-13D5) (all antibodies were kindly provided by Dr J. A. Bluestone), anti-mouse TCR Vγ1 (5 µg/ml, 2.11, BioLegend), or anti-CD40 antibodies (2 µg/ml, IC10, BioLegend). Briefly, Cells (1 × 10⁵ per well) were cultured for 24 h in the presence of medium alone, or plate-bound anti-CD3/TCR mAb or anti-CD40 mAb in 1.0 ml of medium in 96-well plates. After 24 h, the supernatants were harvested and production of IL-4, IFN-γ, IL-17 (R&D Systems) or IL-6 (R&D Systems) measured by ELISA. IL-4 and IFN-γ were detected using capture (ATCC) and biotinylated Abs (BD Biosciences), followed by avidin-HRP then TMB solution (Invitrogen, Carlsbad, CA).

Lung histology

Lung tissue was fixed in Histochoice (AMRESCO, Solon, OH) and embedded in paraffin using a Shandon Citadel tissue processor (Thermo Fisher Scientific, Pittsburg, PA). Microtome sections were cut at 5 µm thickness and stained with H&E using a Shandon Varistain 24–4 (Thermo Fisher Scientific).

Immunohistochemistry

For immunofluorescent staining, lung tissue was frozen in Sakura Tissue-Tek OCT. Cryosections were mounted on glass slides, blocked and dual stained with 100 µl biotinylated anti-IL-17 mAb (10 µg/ml) followed by strepavidin Alexa Fluor 594 (Invitrogen) and 100 µl anti-γδ TCR mAb (UC7-13D5 clone, 200 µg/ml, ATCC) antibodies followed by anti-hamster IgG DyLight 488 (Jackson ImmunoResearch). Coverslips were mounted with FluorSave (EMD, Gibbstown, NJ) and sections were examined using Olympus FV1000 IX81 inverted laser scanning confocal microscope (with spectral detection and TIRF Module). DyLight 488-labelled images (green) were acquired sequentially using the 488 laser line (8% power) and the emission spectra collected at 500–549 nm. Alexa Fluor 594-labelled images (red) were acquired using the 559 laser line (10% power) and emission spectra collected at 589–695 nm. A look up table (LUT) linear adjustment was made for the display.

Pulmonary function measurement

Respiratory resistance (R_L, cm H₂0.s/ml) and dynamic compliance (C_Dyn, ml/cm H₂O) was assessed in anesthetized and tracheotomized mice that were mechanically ventilated in response to increasing concentration of methacholine inhalation (1.5–24 mg/ml) using the pulmonary function equipment from Buxco Research Systems.

Effect of iloprost on IL-17 production

To examine the effect of iloprost on IL-17 production, spleen cells from C57BL/6 mice (2 × 10⁶/ml) were stimulated with soluble anti-mouse γδ TCR (5 µg/ml UC7-13D5, ATTC) in the absence or presence of iloprost (0.5 and 1.0 µM, Cayman Chemical) at 37° C for 4 days. Cell supernatants were then harvested and IL-17 production measured by ELISA (R&D Systems).

Statistical analysis

Data are expressed as means ± SE. Comparisons was analyzed for statistical significance by the Mann Whitney test, with p values <0.05 being considered significant. Analysis was performed with Prism software (GraphPad, La Jolla, CA).

Results

Allergic lung inflammation is associated with an increase in intraepithelial γδ-17 cell numbers

In order to resolve the cellular events involved in allergic lung inflammation that impact on CD4⁺ T cell responses, we utilized both the OVA immunization and the passive CD4⁺ T cell transfer models of asthma. The immunization with OVA is the commonly used method that reproduces many key features of asthma (e.g. airway inflammation, IgE production and AHR), while the adoptive transfer of transgenic Th2 cells approach has the advantage that it allows the tracking of OVA-specific T cells, using the anti-clonotypic TCR antibody KJ1-26, during the inflammatory response. Using the adoptive transfer model allergic airway inflammation, DO11.10 CD4⁺ Th2 cells were generated in vitro and transferred into BALB/c mice that subsequently inhaled aerosolized OVA for 7 consecutive days. Following OVA exposure, Th2 recipients but not control mice developed a pronounced airway inflammation, characterized by a marked increase in the number of lymphocytes and eosinophils and the level of eosinophil peroxidase (EPO) in the BALF (Fig. 1 A). The eosinophilic inflammation was invariably associated with an increase in IL-17-expressing T cells, as well as IL-4-expressing T cells in the lungs (Fig. 1 B). The IL-17-expressing T cells in the lung mononuclear cells (LMC) failed to stain with anti-αβ TCR (Fig. 1 B), anti-clonotypic antibody KJ1-26 (Fig. S1 A) or anti-CD4 (Fig. S1 B) and were thus not of donor origin. Further analysis revealed that the majority of the IL-17-expressing cells were indeed γδ T cells that bore a CD4⁻CD8⁻ phenotype (Fig. S1 B). In sharp contrast, the IL-4-expressing T cells were predominantly αβ T cells and were OVA-specific (KJ1-26⁺, Fig. 1 B). Moreover, LMC from OVA-challenged Th2 recipients produced high levels of IL-17 and IL-4 in response to stimulation with anti-γδ TCR antibody and anti-CD3, respectively (Fig. 1 C). Control mice (no Th2 transfer) did not develop any airway inflammation following OVA inhalation and had low numbers of IL-4 and IL-17-producing T cells present in the lungs (Fig. 1 A, B and C).

FIGURE 1.

Allergic airway inflammation is associated with increased number of γδ-17 cells in the lungs. DO11.10 CD4⁺ Th2 cells were adoptively transferred into BALB/c mice that were then exposed to aerosolized OVA for 7 days. Control mice did not receive Th2 cells but were exposed to OVA aerosols for 7 days. (A) Cell differential counts in the BALF were determined by light microscopic evaluation of stained cytospin preparations. Results are expressed as absolute numbers (per mouse) of lymphocytes (Lym), macrophages (Mac), eosinophils (Eos), and neutrophils (Neu). Eosinophil peroxidase (EPO) levels in the BALF from Th2 recipient or control mice were assessed by colorimetric analysis. (B) The proportion of γδ T cells, αβ T cells and clonotypic KJ1-26⁺ cells co-expressing IL-17 or IL-4 in LMC from Th2 recipient or control mice was determined by FACS analysis using intracellular staining. (C) LMC were stimulated with anti-γδ TCR (5 µg/ml) or anti-CD3 (2 µg/ml) for 24 h and supernatant analyzed for IL-17 or IL-4 production by ELISA. Results are mean ± SE of 4–6 individual mice analyzed per group in triplicates and represent four independent experiments. *p < 0.05.

Since γδ-17 cells require TGF-β for their development (35) and play an important role in orchestrating epithelial barrier function during health and disease (36, 37), we evaluated whether these cells expressed the TGF-β inducible mucosal integrin, αEβ7 (34), during allergic lung inflammation. Interestingly, the majority of IL-17-producing T cells from Th2 recipient mice expressed the αE and β7 integrin chains when characterizing T cells in both the LMC (Fig. 2 A) and BALF (Fig. 2 B). Conversely, this high level of αEβ7 expression was not evident in the control group. Moreover, immunohistological examination of lung tissue revealed expression of IL-17⁺TCR⁺ cells in the airways of Th2 recipients but not control mice (Fig. 2 C). Together, these data show that in tandem with Th2-mediated inflammation there is a marked increase in intraepithelial γδ-17 cells in the airways.

FIGURE 2.

Allergic airway inflammation is associated with increased number of intraepithelial γδ-17 cells. DO11.10 Th2 cells were injected into BALB/c mice that were then exposed to aerosolized OVA for 7 days. Control mice inhaled OVA but did not receive Th2 cells. The proportion of IL-17-expressing cells that co-express γδ TCR, αE integrin or β7 integrin in LMC (A) or BALF (B) from Th2 recipients or control mice was determined by FACS analysis using intracellular staining. (C) Expression of IL-17⁺ and γδ-TCR⁺ cells by lung tissue from Th2 recipient mice or control animals were determined by immunohistochemical staining (20×) and acquired using laser scanning confocal microscopy as described in the Materials and Methods. Data represent four independent experiments with similar results.

γδ-17 cell response in the lung during allergic airway inflammation is dependent on PGI₂

It has been proposed that innate intraepithelial IL-17-producing γδ T cells serve as the sentinels of epithelial surfaces and play a central role in maintaining mucosal barrier integrity. These γδ T cells rapidly produce IL-17 and regulate pathogen clearance, inflammation and epithelial homeostasis in response to tissue stress (12, 38). Given that high levels of PGI₂ are produced during allergic lung inflammation and serve to inhibit the Th2-mediated inflammatory response and remodeling (20, 21, 27), we examined whether this prostanoid exerted any immunoregulatory action on γδ T cell response. IP−/− mice lacking the PGI₂ receptor IP were used and the animals were OVA-immunized and exposed to OVA aerosols for 7 days to induce allergic inflammation. The IP−/− mice had enhanced peribronchial inflammation (Fig. S3 A) with augmented eosinophil numbers and EPO levels in the airways (Fig. S3 B), when compared to OVA-challenged wild-type C57BL/6 (WT) mice. Control IP−/− and WT mice that inhaled PBS did not develop any pulmonary inflammation. In marked contrast to the augmented allergic pulmonary inflammation, a dramatic loss in the proportion (Fig. 3 A) and the absolute number of γδ-17 cells (Fig. 3 B) was observed in the lungs of OVA-challenged IP−/− mice compared to the WT mice. Particularly noteworthy was a loss of γδ T cells expressing αE integrin in the LMC of IP−/− mice (Fig. 3 A). The number of IL-17-expressing αβ T cells was unaffected. Control mice that inhaled PBS had negligible numbers of IL-17-expressing T cells in the lungs (Fig. 3 A). It is important to note that only the γδ-17 cells were affected in IP−/− mice, since the numbers of γδ T cells per se were essentially the same in the lungs of WT and IP−/− mice (5.8% and 5.1%, respectively). It was observed that IL-17-expressing αβ T cells were also present in the lungs of OVA challenged C57BL/6 (Fig. 3 A), and, to a lesser extent in BALB/c mice (Fig. 1 B). This reflects a slightly higher prevalence of “natural” IL-17-expressing αβ T cells in the lungs of naive C57BL/6 compared to BALB/c mice (Fig. S2). These cells were found to be CD4⁻CD8⁻ iNKT cells typically present in both the lungs and spleen (data not shown), and have been described previously (39).

FIGURE 3.

Loss of γδ-17 cells in the lungs of IP−/− mice during allergic inflammation. C57BL/6 WT or IP−/− (IPKO) mice were immunized with OVA and after 10 days the mice were exposed to either aerosolized OVA or PBS (control) for 7 days. (A) The proportion of γδ T cells and αβ T cells co-expressing IL-17 or αE integrin in the LMC obtained from OVA-challenged or control IP−/− and WT mice determined by FACS analysis using intracellular staining. (B) The absolute number/mouse of γδ T cells expressing IL-17 in LMC of OVA-challenged or control WT and IP−/− mice. Data are mean ± SE (n=3). FACS data represent four independent experiments. *p < 0.05, compared to OVA challenged WT.

Consistently, there was a pronounced loss of IL-17 production by γδ T cells present in both the LMC and spleens (stimulated with anti-γδ TCR or anti-Vγ4 antibody) of IP−/− mice compared to WT mice (Fig. 4 A). This loss of IL-17 production by γδ T cells appears to comprise mainly of Vγ4⁺ cells (50–60% of the IL-17 production by WT γδ T cells was by Vγ4⁺ cells, Fig. 4 A) which is in accordance with the report of Murdoch et al (40). Consistent with this data, typically, 30% of the γδ T cells present in the LMC of OVA-challenged WT and IP−/− mice were Vγ4⁺ T cells (Fig. 4 B). Stimulation of LMC with plate-bound anti-Vγ5 or anti-Vγ1 antibodies showed that the remaining IL-17 production was not by Vγ5⁺ or Vγ1⁺ T cells (data not shown). In keeping with these findings, mice receiving Th2 cells and treated daily with indomethacin, a non-selective cyclooxygenase 1 and 2 inhibitor, elicited a reduced γδ-17 cell response in the lungs (Fig. 4 C). Collectively, these results demonstrate that prostanoids, and PGI₂ in particular, plays a key role in promoting γδ-17 T cell response in the lung.

Figure 4. Loss of γδ-17 cells in the lungs of IP−/− and indomethacin-treated mice: role for Vγ4 γδ T cells.

(A) C57BL/6 WT or IP−/− (IPKO) mice were immunized with OVA and after 10 days the mice were exposed to either aerosolized OVA or PBS for 7 days. LMC or splenic cells (SPLN) were harvested, stimulated for 24 h with anti-γδ TCR or anti-Vγ4 (5 µg/ml) and IL-17 production in the supernatants was then measured by ELISA. (B) The proportion of γδ T cells expressing Vγ4 in LMC from OVA challenged WT and IPKO mice analyzed by FACS. (C) LMC from naïve WT or IPKO mice were stimulated for 24 h with anti-CD3 (2 µg/ml) and IL-17 production determined by ELISA. (D) DO11.10 CD4⁺ Th2 cells were adoptively transferred into BALB/c mice that were then either treated daily by ip injection of 5 mg/kg of indomethacin (Th2 + INDO group) or PBS vehicle (Th2 group). Control mice did not receive Th2 cells (Con). All mice were exposed to OVA aerosols for 7 days. Results are mean ± SE of 4–6 individual mice analyzed per group in triplicates and represent three independent experiments. *p < 0.05 compared with WT or Th2 group.

Loss of programming of γδ-17 cells in the thymus and spleen of naïve IP−/− mice: enhancement by iloprost

We next investigated whether this loss of γδ T cells was a consequence of a preexisting defect in the generation of natural γδ-17 cells in the thymus and spleen of IP−/− mice. We discovered a dramatic reduction in the proportion (Fig. 5 A) and total number (Fig. 5 B) of natural γδ-17 cells in both the thymus and spleen from naïve IP−/− mice. Similar to lung γδ T cells, the majority of natural γδ T cells in the thymus expressed αEβ7 integrin (Fig. 5 A). As depicted in Fig. 5B, the absolute number of γδ-17 cells per mouse was 8.5 × 10⁵ in spleen of naïve WT mice, compared to 4.1 × 10⁵ in IP−/− mice. In marked contrast, the number and proportion of IL-17⁻ γδ TCR⁺ cells was similar in both IP−/− and WT tissues, suggesting that the absence of PGI₂-IP signaling affects IL-17 expression rather than favoring the expression of particular γδ TCR gene rearrangements.

FIGURE 5.

Loss of γδ-17 cells in the spleen and thymus of naïve IP−/− mice: enhancement by iloprost. (A) The proportion and (B) the absolute number/mouse of γδ T cells co-expressing IL-17, αE integrin or β7 integrin in the spleen (depleted of CD4⁺, CD8⁺ and B220⁺ cells) or thymus (depleted of CD4⁺ and CD8⁺ cells) from naïve C57BL/6 WT mice or IP−/− mice was determined by FACS analysis using intracellular staining. (C) IL-17 production by splenic γδ T cells stimulated with anti-γδ TCR in the absence or presence of iloprost, a stable analog of PGI₂. Spleen cells (2 × 10⁶/ml) from naïve C57BL/6 mice were stimulated with soluble 5 µg/ml anti-γδ TCR ± iloprost (0, 0.5 and 1.0 µM) for 4 days and IL-17 production measured by ELISA. Data are mean ± SE (n=3). FACS results represent three to four independent experiments. *p < 0.05.

Conversely, iloprost (1.0 µM), a stable analog of PGI₂, significantly enhanced IL-17 production by splenic γδ T cells stimulated with anti-γδ TCR antibody (5 µg/ml) (Fig. 5 C). Relatively low levels of IFN-γ and IL-4 were produced by the γδ T cells (data not shown). In summary, a pronounced reduction in the number of γδ-17 cells (but not γδ T cell numbers) was observed in the thymus and spleen of naïve IP−/− mice, implying a critical role for PGI₂ in the programming of natural γδ-17 cells.

IL-6 production by eosinophils and dendritic cells during allergic lung inflammation is dependent on PGI₂ and promotes γδ-17 cell development

Given that IL-6 is required for promoting the development of IL-17-producing αβ-TCR expressing T cells (5), we examined whether PGI₂ facilitated IL-6 production and subsequent γδ-17 cell development. In the first instance we examined the role of IL-6 in generating the γδ-17 response. Using the Th2 transfer model of asthma, treatment of Th2 recipient mice with intranasal anti-IL-6 mAb (0.6 mg/kg, every 48 h during the 7 day OVA inhalation period) caused a reduction in the number of γδ-17 cells and CD103⁺ cells (αE-expressing), but not αβ T cells, compared to vehicle-treated Th2 recipient mice (Fig. 6 A). In sharp contrast, OVA-challenged control animals had negligible numbers of IL-17-expressing T cells in the lungs.

FIGURE 6.

IL-6 production by eosinophils during allergic pulmonary inflammation is dependent on PGI₂-IP signaling and promotes γδ-17 cell development. (A) DO11.10 CD4⁺ Th2 cells were transferred into BALB/c mice that were either treated with intranasal anti-IL-6 mAb (Th2 + anti-IL-6, 0.6 mg/kg in 30 µl PBS) or PBS vehicle (Th2) four times every 48 h during the 7 day OVA inhalation period. Control mice were vehicle-treated but did not receive Th2 cells. All mice were exposed to aerosolized OVA for 7 days. IL-17-expressing αβ T cells, γδ T cells and CD103⁺ (αE⁺) cells in the LMC was determined by FACS analysis and expressed as absolute cell number per mouse. (B) C57BL/6 WT or IP−/− (IPKO) mice were immunized with OVA and after 10 days the mice were exposed to either aerosolized OVA or PBS (control) for 7 days. Number of lung CD11b⁺CCR3⁺ eosinophils co-expressing IL-6 (expressed as mean fluorescent intensity (MFI)) and expression of IL-6 by whole LMC from challenged IPKO and WT or control mice was determined by FACS analysis using intracellular staining. Results are mean ± SE (n=4) and represent three to four independent experiments. *p < 0.05 compared with vehicle-treated Th2 group.

To examine the cellular source of IL-6 in the lung during allergic inflammation, C57BL/6 WT or IP−/− mice were immunized with OVA/Alum prior to exposure to either aerosolized OVA or PBS (control) for 7 consecutive days. The cellular source of IL-6 production during allergic airway inflammation was identified by intracellular staining using three-color flow cytometry. IL-6 expression by CD11b⁺, CD11c⁺, Class II⁺, Gr-1⁺ or CCR3⁺ cells in LMC from OVA-challenged or control animals was determined in this way. By gating on forward and side-scatter, the majority IL-6-expressing cells in the lungs during allergic inflammation were found to comprise of CD11b⁺CCR3⁺ eosinophils and to a lesser extent CD11c⁺ Class II⁺ dendritic cells (Fig. 6 B). Moreover, a marked increase in the number of CD11b⁺CCR3⁺ eosinophils were found in the lungs of both OVA-challenged WT (15.1%) and OVA-challenged IP−/− (18.5%) mice compared to control animals (3.1%) (Fig. 6 B). Importantly, the level of IL-6 expression by the total LMC population and, in particular, the CD11b⁺CCR3⁺ eosinophils were markedly less in IP−/− compared to WT mice (mean fluorescence intensity (MFI) of 800 vs 1403, respectively) (Fig. 6 B). It is important to note that the CD11b⁺CCR3⁺ cells were initially sorted using FACSAria II and cytospin preparations were stained with Hema3 to confirm eosinophil purity by light microscopic evaluation which was > 98%. IL-6-expression by lung CD11c⁺ Class II⁺ dendritic cells was also less in OVA-challenged IP−/− compared to WT mice (MFI of 1106 vs 1330, respectively).

Collectively, these results suggest that the PGI₂-dependent IL-6 production by eosinophils and dendritic cells promotes the development of intraepithelial γδ-17 cells.

Iloprost treatment reduces allergic airway inflammation and AHR but enhances IL-6 production

To examine the immunomodulatory effects of PGI₂ in vivo, mice were treated with the stable analog of PGI₂, iloprost, and the effect on allergic airway inflammation and AHR was determined. Using the Th2 transfer model of asthma, Th2 recipient mice were intranasally administered either with iloprost (0.1 mg/kg) or vehicle, every 48 h during the OVA inhalation period. Control mice did not receive any Th2 cells but were vehicle-treated. All mice were exposed to aerosolized OVA for 7 consecutive days. Following OVA inhalation, there was a pronounced influx of lymphocytes and eosinophils into the airways of Th2 recipients (Fig. 7 A) with increased EPO levels in their BALF compared to control animals (Fig. 7 B). Treatment with iloprost significantly reduced the number of eosinophil and lymphocyte accumulation in the airways (Fig. 7 A, B), but caused a two-fold increase in IL-6 production by LMC (Fig. 7 C). Moreover, there was approximately 50% inhibition in the proportion (Fig. 7 D) and total number (Fig. 7 E) of OVA-specific CD4⁺KJ1-26⁺ T cells in the lungs of mice treated with iloprost. Control animals had negligible levels of CD4⁺KJ1-26⁺ T cells in the lungs (Fig. 7, D and E). Consistently, an augmented level of AHR, depicted by increased resistance and decreased compliance, was observed in the Th2 recipient compared to control mice which was reduced by iloprost treatment (Fig. 7 F).

FIGURE 7.

Iloprost treatment downregulates allergic airway inflammation and AHR but enhances IL-6 production. DO11.10 CD4⁺ Th2 cells (6 × 10⁶/mouse) were adoptively transferred into BALB/c mice and were either treated with intranasal iloprost (Th2 + ILOP, 0.1 mg/kg in 30 µl PBS) or PBS vehicle (Th2) every 48 h during the 7 day OVA inhalation period (i.e. 4 instillations). Control mice were vehicle-treated but did not receive Th2 cells. All mice were exposed to aerosolized OVA for 7 days. (A) Cell differential counts in the BALF were determined by light microscopic evaluation of stained cytospin preparations. Results are expressed as absolute numbers (per mouse) of lymphocytes (Lym), macrophages (Mac), eosinophils (Eos), and neutrophils (Neu). (B) The level of eosinophil peroxidase (EPO) levels in the BALF was assessed by colorimetric analysis. (C) IL-6 production by LMC from Th2 or Th2 + ILOP group was determined by ELISA. The proportion (D) and absolute number (E) of clonotypic CD4⁺KJ1-26⁺ T cells in LMC was determined by FACS analysis. (F) Lung resistance (R_L) and dynamic compliance (C_Dyn) was assessed in anesthetized and tracheotomized mice that were mechanically ventilated in response to increasing concentration of methacholine inhalation. Data are mean ± SE of 4–6 individual mice analyzed per group in triplicates and represent three to four independent experiments. *p < 0.05 compared with vehicle-treated Th2 group.

Depletion of Vγ4 γδ T cells in vivo augments airway Th2-mediated inflammation

To investigate the contribution of Vγ4 γδ T cells in the allergic inflammatory process, anti-Vγ4 antibody (1 mg/kg) was given intranasally to mice in order to deplete Vγ4 γδ T cells that have accumulated in the airways. BALB/c mice were first adoptively transferred with DO11.10 CD4⁺ Th2 cells and the mice were treated intranasally with either anti-Vγ4 antibody (Th2 + anti-Vγ4) or vehicle (Th2) every 48 h during the OVA inhalation period. Control mice were vehicle-treated but did not receive any Th2 cells (no transfer). All mice inhaled aerosolized OVA for 7 consecutive days. Treatment of Th2 recipient mice with the anti-Vγ4 antibody caused a marked decrease in the proportion (Fig. 8 A) and number (Fig. 8 B) of CD103⁺γδ⁺ T cells in the lungs compared to Th2 recipients untreated with the antibody. This depletion of intraepithelial Vγ4 γδ T cells resulted in an augmented number of antigen-specific T cells (CD4⁺KJ1-26⁺ cells) and an increase in the number of eosinophils and the level of EPO activity in the airways (Fig. 8, C, D and E). Control mice (no Th2 transfer) did not develop any airway inflammation. Together, these results suggest that Vγ4 γδ T cells play an important immunoregulatory role during allergic pulmonary inflammation.

FIGURE 8.

Vγ4 depletion in vivo augments Th2-mediated pulmonary inflammation. DO11.10 CD4⁺ Th2 cells were adoptively transferred into BALB/c mice and were administered intranasally (every 48 h during 7 day OVA inhalation period) with either anti-Vγ4 mAb (Th2 + anti-Vγ4, 25 µg in 30 µl PBS) or PBS vehicle (Th2). Control mice were vehicle-treated but did not receive Th2 cells (no transfer). All mice were exposed to aerosolized OVA for 7 c days. The proportion (A) and absolute number (B) of γδ T cells, co-expressing αE integrin (CD103) in LMC from Th2, Th2 + anti-Vγ4, or no transfer groups were determined by FACS analysis. (C) The number of CD4⁺KJ1-26⁺ T cells was determined by FACS analysis, and results expressed as cell number per mouse. (D) Cell differential counts in the BALF were determined by light microscopic evaluation of stained cytospin preparations. Results are expressed as absolute numbers (per mouse) of lymphocytes (Lym), macrophages (Mac), eosinophils (Eos), and neutrophils (Neu). (E) EPO levels in the BALF were assessed by colorimetric analysis. FACS results are representative of two independent experiments. Data are mean ± SE of 4–6 individual mice analyzed per group in triplicates and represent three independent experiments. *p < 0.05 compared with Th2 group.

Discussion

Using a Th2 adoptive transfer model of allergic lung inflammation, we have previously examined the CD4⁺ Th2 response and its regulation by the prostanoid PGI₂ generated during the inflammatory response (20, 21). Curiously, we observed that during the allergic inflammation, IL-17-producing T cells accumulated in the airways. In the present study, we sought to use this model to characterize these IL-17-expressing T cells. Surprisingly, the IL-17-producing T cells in the inflamed lungs were predominantly γδ T cells. Although only low numbers of γδ T cells were found to be resident in the lung tissue of naïve mice, following the onset of Th2-mediated airway eosinophilic inflammation, a marked increase in the number of host intraepithelial CD4⁻CD8⁻ γδ T cells in the lungs was noted. Moreover, the vast majority of γδ T cells in this inflammatory site produced IL-17. The accumulation of γδ-17 cells in the lung during mucosal inflammation induced by inhaled allergen was intriguing and prompted speculation that they may play a role in the inflammatory process or its regulation. Strikingly, the γδ T cells in the inflamed lung tissue uniformly expressed the αEβ7 integrin that promotes adhesion to E-cadherin (32) and, expectedly, these cells were largely associated with the airway epithelium. Such “priming” of the airway epithelium with γδ-17 cells during allergic inflammation is consistent with the proposed key function of these cells as sentinels of epithelial surfaces (12). However, many questions remain that pertain to the development of these cells in the thymus and periphery, the nature of antigen recognized and their role in mucosal inflammation. Certainly, the juxtaposition of these cells to the epithelium is strongly suggestive of them playing a role in modulation of the behavior of airway epithelial cells during the inflammatory phase.

Our previous studies have revealed that PGI₂ limits the progression of CD4⁺ Th2 cell responses (20, 21) primarily since the IP receptor for PGI₂ is upregulated by IL-4 generated during allergic lung inflammation. As a result of this, the immunoregulatory properties of this prostanoid are most evident during Th2-mediated inflammatory responses. Consequently, we examined whether the γδ T cell response was also influenced by PGI₂ by using mice lacking the IP receptor. Our data revealed that allergic lung inflammation was augmented in IP−/− mice (as previously shown (19)) but, in stark contrast, the appearance of γδ-17 cells in the lungs of these animals was attenuated. This was surprising since the emergence of γδ-17 cells closely paralleled the level of allergic inflammation. Consequently, this observation strongly suggested that PGI₂ is an essential component, underpinning the lung γδ-17 cell response. This effect stemmed from a markedly reduced number of “natural” innate γδ-17 cells in the IP null mice. This defect was also evident in the thymus where a failure to generate γδ-17 cells expressing the αEβ7 integrin was noted in naïve IP−/− mice. Conversely, the stable analog of PGI₂, iloprost, markedly increased the IL-17 production by splenic γδ T cells but significantly reduced the airway inflammation.

The pronounced reduction in IL-17 production by γδ T cells evident in IP−/− mice was surprising and strongly implied that PGI₂ played a critical role in the programming of IL-17 production by these cells in the thymus, and possibly in the periphery. This defect in the IP−/− mouse could not be a consequence of altered γδ TCR expression per se since their total numbers and Vγ usage were similar to WT littermates. To date, both TGF-β and RORγt have been shown to be essential for the generation of natural γδ-17 cells (35, 41). Furthermore a role for IL-23 in promoting IL-17 release by γδ T cells has been proposed (42, 43) and the augmentation of IL-17 production by αβ T cells by PGE₂ by an IL-23-dependent mechanism has been well documented (44–46). In contrast, PGI₂ and its receptor played an important role in augmenting IL-6 production by eosinophils, and also by dendritic cells, which have been shown to express IP (47). That eosinophils are responsive to PGI₂ might be expected from our previous finding that IL-4 is an important cytokine for inducing expression of the receptor (21) and reports that eosinophils are an important source of this cytokine which was clearly illustrated in mouse eosinophils using IL-4 GFP reporter mice (48). Certainly, human eosinophils impact on the inflammatory process by releasing a range of cytokines that include IL-4, IL-13, IL-6, TGF-β and IL-10 (49). Conceivably, during allergic inflammation the programming of cytokine expression is strongly influenced by PGI₂ in an environment where IL-4 plays a central role. The observation that PGI₂ played an important role in both the development of natural and inflammatory γδ-17 cells is indicative of this mechanism being operative both during the generation of these cells in the thymus and in the periphery during allergic inflammation, although the relative importance of eosinophils and dendritic cells may differ in these two scenarios.

γδ T cells differ markedly from αβ T cells in their TCR receptor diversity and a propensity to localize to epithelial sites (50). The preferential homing of γδ T cells to epithelial tissues is an intrinsic function of this cell type, exemplified by the observation that approximately one third of the intestinal intraepithelial cell express a γδ TCR (50). The use of a particular Vγ segment by the γδ TCR is highly relevant since in early life the diversity of the γδ T cell receptor is a function of embryonical stage of development, with T cells, generated early and expressing a canonical γδ TCR using Vγ5 and Vγ6, emigrating from the thymus to the skin and female reproductive tract, respectively (51, 52). Subsequently, T cells leaving the thymus display higher levels of diversity and seed into peripheral sites. Interestingly, in the present study a significant amount of IL-17 production in the lung was found by Vγ4⁺ γδ T cells but not Vγ5 cells or Vγ1. Vγ4-expressing cells in the lung have been observed previously during OVA-induced lung inflammation (13, 14, 53) or respiratory syncytial virus infection (54). It has been demonstrated that both short and long-term OVA inhalation induced Vγ4 expressing suppressor cells that inhibited AHR (53, 55, 56) and reduced the IgE response (57). This Vγ4 response required CD8⁺ dendritic cells in order to develop (55) and did not affect the inflammatory response (14). Vγ6⁺ cells have been shown to be present in the lungs of mice following infection with Bacillus subtilis (58), while Vγ1⁺ cells promote AHR in a model of allergic inflammation (59). More recent work has shown that of IL-17-producing γδ T cells are involved in the resolution of allergic airway inflammation and AHR (40).

The priming of the airways with intraepithelial γδ-17 cells raises the issue as to whether this contributes to the inflammatory process. A notable property of γδ T cells is the rapid cytokine release on encountering antigen, a characteristic that is attributed to the prior programming of these T cells in the thymus and possible polarization in the periphery (38). As such, these cells are regarded as “first-responders” and provide an immediate response to environmental insult or infection. γδ T cells are able to recognize non-peptidic antigens expressed by stressed cells recognized by pattern recognition receptors (PPR). However, innate responses elicited by γδ T cells can also ensue following engagement of the PPR Dectin-1 and Toll-like receptor 2 (42, 43). γδ T cells that produce IL-17 have been implicated in a series of immune responses which include the clearance of Mycobactrium tuberculosis (60), Escherichia coli (61, 62) and Candida albicans (63). IL-17 production in the mucosal site is of particular importance since this cytokine exerts a wide range of effects at epithelial surfaces that include release of β-defensins (64), promoting the recruitment of neutrophils (65), inducing the expression of the polymeric immunoglobulin receptor and trans-epithelial transport of IgA (9, 10). Conceivably, γδ-17 cells play a role in the recruitment of neutrophils to the lung during allergic lung inflammation. Although the most striking aspect of allergic lung inflammation is the pronounced infiltration of eosinophils into the airways several reports have documented the involvement of neutrophils in severe human asthma (66–68). Given the pronounced immunoregulatory properties that have been attributed to Vγ4 T cells, we examined their contribution to allergic lung inflammation using the adoptive transfer model of asthma. Our results show that the depletion of the Vγ4⁺ subset by intranasal administration of anti-Vγ4 antibody was associated with an increase in the recruitment of antigen-specific CD4⁺ Th2 cells into the lung following OVA inhalation. This effect was coincident with an increase in the number of eosinophils present in the airway. Collectively, these findings strongly suggest that these intraepithelial γδ T cells limit the eosinophilic inflammation by attenuating the response of CD4⁺ Th2 cells to the inhaled allergen. Although it has been reported that Vγ4 expressing T cells downregulate AHR but exert a minimal effect on the inflammation (14, 59), this difference may arise because of the passive immunization used in our study obviated the use of an alum adjuvant, and the induction of characteristically different T cell response.

A major question posed by these findings pertains to the biological significance of the observation that the development of γδ-17 T cells is critically dependent on PGI₂. In marked contrast to their αβ counterparts, γδ T cells typically recognize a range of non-peptidic antigens. In humans Vγ9Vδ2⁺ T cells are activated by phosphoantigens that are produced by the isoprenoid pathway (69) and a high proportion of peripheral blood γδ T cells are activated by small phosphorylated (70) or aminated alkyl molecules (71). γδ T cells in both humans and mice have been shown to recognize phospholipids (72–74)) and CD1d associated lipids (75). PGI₂ is not only a product of lipid metabolism but also reduces the liberation of arachidonic acid from membranes phospholipid by inhibiting PLA₂ production (76). In this respect, the recognition of phospholipids by the γδ T cell receptor is relevant since prostacyclin biosynthesis requires the release of arachidonic acid from phospholipid stores.

In summary, we demonstrated that the allergic airway inflammatory response was associated with a marked increase in the number of innate intraepithelial γδ-17 cells and, importantly, that PGI₂-IP signaling is critical for the development of these cells through an IL-6-dependent mechanism. That the adaptive immune system elicits an epithelial-associated γδ-17 T cell response raises the possibility that such epithelial-tropic cells play an important local immunoregulatory role. These findings reveal molecular mechanisms that are likely to be operative at mucosal sites and point to a role for prostanoids in molding the developing immune response. Conceivably, modifying intraepithelial γδ T cell function using stable PGI₂ analogs, such as iloprost, may provide a novel approach capable of regulating some of the changes to the airway epithelium that occur in chronic lung inflammatory diseases such as asthma.

Abbreviations used in this article

AHR

airway hyperreactivity

BALF

bronchoalveolar lavage fluid

EPO

eosinophil peroxidase

γδ-17 cells

IL-17-producing γδ T cells

LMC

lung mononuclear cells

LN

lymph node

PG

prostaglandin