IL-17A–Producing γδ T and Th17 Lymphocytes Mediate Lung Inflammation but Not Fibrosis in Experimental Silicosis Free

Tissue injury causes accumulation and activation of neutrophils in the lung. The persistence of tissue damage, abnormal repair, and, finally, fibrosis lead to a variety of chronic pulmonary diseases. Experimental silicosis represents a useful model to explore the mechanisms and mediators involved in these reactions (1, 2).

Cytokines and chemokines have been identified as powerful mediators recruiting neutrophils into pulmonary tissue after silica particle exposure (3). The uptake of silica particles by macrophages triggers the production of IL-1β by activation of the apoptosis associated speck-like protein- and NALP3-dependent inflammasome (4–6). Reactive oxygen intermediates produced by silica particles and/or by activated macrophages regulate TNF-α production (7). These proinflammatory cytokines are crucial for neutrophil influx and activation (8). In addition to macrophages, alveolar epithelial cells secrete high levels of CXC chemokines, leading to massive neutrophil influx upon silica exposure (9). The role of other immune cells, such as Th17 lymphocytes, in these reactions has not been explored.

Th17 lymphocytes represent a recently identified Th cell lineage that has a proinflammatory role in autoimmunity and neutrophilic inflammation. The differentiation of Th17 cells requires TGF-β, IL-6, and/or IL-23. The transcription factors STAT3 and retinoic acid-receptor–related orphan receptor-γ mediate their lineage commitment. Th17 cells participate in neutrophil accumulation because they are an important source of IL-17A and -22, which regulate CXC chemokine and G-CSF production necessary for neutrophil differentiation, activation, and recruitment (10–12). CD4⁺ T cells are not the only source of IL-17 and -22; γδ T lymphocytes also produce considerable quantities of these cytokines in different neutrophilic reactions. NK T cells and CD8 T cells were also identified as IL-17–producing cells that contribute to neutrophil recruitment through IL-17 secretion (13–15).

In this article, we report an accumulation of IL-17A– and -22–producing T lymphocytes in experimental silicosis. To investigate their role during inflammation and fibrosis, we used IL-17A–neutralizing Abs and IL-17R– and IL-22–deficient mice and assessed the effects of recombinant mouse IL-17A and -22 on fibroblast functions. We demonstrate that the early lung neutrophilic inflammation and acute tissue injury require the production of IL-17A by γδ T lymphocytes and Th17 cells. However, these lymphocytes and their related cytokines are dispensable for the establishment of late chronic inflammation and lung fibrosis.

Mice were bred and maintained under specific pathogen-free conditions at the Université catholique de Louvain. The protocol of this investigation was approved by the local committee for animal research. Crystalline silica particles (DQ12; d₅₀ = 2.2 μm, a gift from Dr. Ambruster, Essen, Germany) were suspended in sterile 0.9% saline (Braun Medical, Diegem, Belgium) at 50 mg/ml and administered by pharyngeal instillation at a dose of 2.5 mg/mouse. To sterilize and inactivate any trace of endotoxins, particles were heated to 200°C for 2 h immediately before suspension and administration. Female IL-17Rα–deficient (IL-17R^−/−) (16), γδ T lymphocyte–deficient (TCRγδ^−/−) (17), IL-23 p19–deficient (IL-23^−/−) (18), and IL-22–deficient (IL-22^−/−) (19) mice, all in a C57BL/6 background, were bred locally; NMRI nude and wild-type mice were purchased from Taconic Europe (Bomholt, Denmark). BALB/c, C57BL/6J, and DBA/2 mice were obtained from our local breeding facility (G. Warnier, Ludwig Institute for Cancer Research, Brussels, Belgium).

One day before and after the instillation of saline (controls) or silica, C57BL/6 mice were injected i.p. with mouse anti–mIL-17A IgG1 (MM17F3; 0.5 mg/100 μl PBS) or rat anti–mIL-17A IgG2a (MAB421 clone 50104; 50 μg/100 μl PBS, R&D Systems, Minneapolis, MN). Anti–IL-17A mAb (MM17F3) was also administered twice a week for 60 d in long-term studies. Rat anti-mouse IL-23p40 IgG2a (C17.8; 250 μg/100 μl PBS), rat anti-mouse CD4 IgG2b (L3T4; clone GK 1.5, kindly provided by J.P. Coutelier, Experimental Medicine Unit, Université catholique de Louvain), rat anti-mouse IL-6R IgG2b (clone D7715A7; 1 mg/100 μl PBS), and rat anti-mouse IL-6 IgG1 (clone D6906B4, 1 mg/100 μl PBS) were administered i.p. on days −1 and +1 after silica. For local blockade of TGF-β, mouse anti–TGF-β1, 2, and 3 IgG1 (clone 1D11; 100 μg/50 μl PBS) were pharyngeally injected also at days −1 and +1. The Abs were purified according to standard procedures, and all dosing regimens were previously found to be sufficient for neutralization of IL-17A (20–22), IL-23p40 (23), CD4 (24), IL-6/IL-6R (25), and TGF-β (26). IgG isotype controls were mouse anti-trinitrophenyl IgG1 (B9603D11), mouse anti-keyhole limpet hemocyanin IgG1 (MAB002; clone 11711, R&D Systems), and rat anti-keyhole limpet hemocyanin IgG2a (MAB006; clone 54447, R&D Systems).

Liposomes with or without clodronate were produced as previously described (27–29). Alveolar macrophage depletion was achieved by the pharyngeal instillation of 100 μl liposome-encapsulated clodronate suspension (0.5 mg) 24 h after silica or saline treatment. Macrophages in bronchoalveolar lavage (BAL) were depleted by >80% with this regimen, as assessed at day +3 after saline or silica. Control mice received empty (saline-containing) liposomes.

At selected time-points following silica instillation, mice were sacrificed with an i.p. injection of sodium pentobarbital (20 mg/mice), and BAL was performed by cannulating the trachea and lavaging the lung four times with 1 ml NaCl 0.9%. The BAL fluid (BALF) was centrifuged at 280 × g, for 10 min at 4°C (Centrifuge 5804R; Eppendorf, Hamburg, Germany), and the cell-free supernatant was used for biochemical measurements (see below). Aliquots of cell pellets were used to determine total cell numbers and differentials. The cells were cytocentrifuged onto glass slides and subjected to Diff-Quick staining (Dade Behring, Deerfield, IL). Polymorphonuclear and mononuclear cells were counted by light microscopy at ×200 magnification (total of 300 cells counted). Lactate dehydrogenase activity in BALF was assayed spectrophotometrically by monitoring the reduction of NAD (NAD⁺) (Sigma-Aldrich, St. Louis, MO) at 340 nm in the presence of lactate. Total protein in BALF was determined by the pyrogallol red staining method (Technicon RA system; Bayer Diagnostics, Domont, France). Lung tissue myeloperoxidase (MPO) activity was estimated in lung homogenates, as described (30). Murine IL-17A, -6, and -1β, TNF-α, keratinocyte-derived chemokine (KC), MIP-2, and TGF-β1 concentrations were measured by ELISA kits (R&D Systems), following the manufacturer’s protocols.

The lavaged lungs were perfused with NaCl 0.9% through the right ventricle, excised, and frozen at −80°C. The lungs were then homogenized for 1 min on ice with an Ultra-Turrax T25 homogenizer in 2 ml cold PBS (Janke & Kunkel, Brussels, Belgium). Hydroxyproline (OH-proline) was assessed by HPLC on hydrolyzed lung homogenates (6 N HCl at 108°C for 24 h), as previously described (31). Soluble collagen levels in lung homogenate supernatants were assessed by the Sircol collagen assay, according to the manufacturer’s protocol (Biocolor, Belfast, Ireland). Fibronectin and type I collagen contents in lungs and fibroblast culture supernatants were measured by ELISA, as previously described (32).

Macrophages, granulocytes, and lymphocytes were isolated from the pulmonary cell suspension obtained by lung digestion (32). After centrifugation, macrophages, neutrophils, and γδ⁺ and CD4⁺ T lymphocytes were separated and isolated using immunomagnetic beads and a magnetic cell-separation system (Miltenyi Biotec, Auburn, CA). The cells were purified by positive selection using anti-CD11c, anti-GR1, anti-TCRγδ, or anti-CD4 magnetic beads, respectively (Miltenyi Biotec). The purity of immune cell preparations was routinely >95%, as assessed by Diff-Quick staining.

Lung cells were isolated as mentioned above and stimulated for 4 h with PMA (20 ng/μl; Sigma-Aldrich), ionomycin (100 ng/μl; Sigma-Aldrich), and monensin (BD GolgiStop protein transport inhibitor, BD Biosciences, Erembodegem, Belgium and Luxembourg). FcRs were blocked with anti-CD16/32 (clone 2.4G2; BD Biosciences). Cells were stained using Abs specific for CD5 (clone Ly-1), CD4 (clone GK1.5), and TCRγδ (clone GL3), followed by incubation with the Fixation/Permeabilization solution (BD Cytofix/Cytoperm kit; all from BD Biosciences). Cells were then stained for intracellular IL-17A (21), washed with the BD Perm/Wash buffer, and fixed in a 1.25% paraformaldehyde solution in PBS for ≥1 h. Cells were acquired on a FACSCalibur and analyzed using the CellQuest software (both from BD Biosciences). Analysis of cell populations was done with appropriate gating, according to side and forward scatters, to exclude dead cells and silica particles.

BAL cell pellets, MACS-purified BAL cells, and nonlavaged lungs were homogenized on ice in 2 ml TRIzol reagent (Invitrogen, Merelbeke, Belgium), and total RNA was extracted according to the manufacturer’s instructions. Residual DNA contamination was removed by treatment with DNAse-free (Ambion, Austin, TX). One microgram of RNA was reverse transcribed with Superscript RNase H-reverse transcriptase (Invitrogen) with 350 pmol random hexamers (Eurogentec, Seraing, Belgium). The resulting cDNA was diluted 25× or 50× and used as template in subsequent PCR assays.

Sequences of interest were amplified by PCR using the following forward primers (Invitrogen): 5′-AGA GGG AAA TCG TGC GTG AC-3′ (mouse β-actin), 5′-GCC TCT TCT CAT TCC TGC TTG-3′ (mouse IL-17A), and 5′-CGC CTA CCA CAT CCA AGG AA-3′ (18S RNA) and reverse primers: 5′-CAA TAG TGA TGA CCT GGC CGT-3′ (mouse β-actin), 5′-GGC CAT TTG GGA ACT TCT CA-3′ (mouse IL-17A), and 5′-GGT AAT TCC AGC TCC AAT AGC GTAT-3′ (18S RNA). For IL-22 and -23p19, TaqMan predesigned primers were used (TaqMan gene-expression assay, Applied Biosystems, Foster City, CA). PCR was performed with AmpliTaq Gold polymerase (Invitrogen, Merelbeke, Belgium), according to the manufacturer’s instructions, with the following temperature program: 2 min at 94°C; 30 s at 94°C, 30 s at 55°C, and 20 s at 72°C for 40 cycles; and 5 min at 72°C. Amplified DNA fragments were purified from a 1.5% agarose gel with the CR clean-up, Gel extraction kit (Macherey-Nagel, Düren, Germany) and then serially diluted to serve as standards in real-time PCR. Reverse-transcribed mRNAs were finally quantified by real-time PCR using SYBR Green technology on an ABI Prism 7000 Sequence Detection System (Applied Biosystems), according to the following program: 2 min at 50°C, 10 min at 95°C, and 15 s at 95°C and 1 min at 60°C for 40 cycles. To verify specific amplification, a dissociation curve was obtained by increasing the temperature to 95°C for 20 min. For IL-22, -17F, and -23p19, the real-time PCR was performed using TaqMan technology on an ABI Prism 7000 Sequence Detection System (Applied Biosystems), according to the following program: 2 min at 50°C, 10 min at 95°C, and 15 s at 95°C and 1 min at 60°C for 40 cycles. Five microliters of diluted cDNA or standards were amplified with 300 nM the appropriate primers using SYBR Green PCR Master Mix or with 900 nM the IL-22, -17F, or -23p19 primers using TaqMan Universal Master Mix (both from Applied Biosystems), in a total volume of 25 μl. Results were calculated as a ratio of gene expression to the expression of the reference gene, β-actin or 18S RNA.

Mouse lung fibroblasts were isolated from lung tissue by mincing and enzymatic digestion, as previously described (33). Cells were cultured in complete medium composed of DMEM supplemented with 10% FBS and antibiotics (Invitrogen). Fibroblasts used after the first passage were seeded into 24- or 96-well plates at 40 × 10³ and 10 × 10³ cells/well, respectively. Subconfluent monolayers were treated for 24 h with various concentrations of recombinant mouse IL-17A (R&D Systems) or IL-22 (34) suspended in medium supplemented with 0.1% FBS. Fibroblast proliferation was estimated by [³H]thymidine incorporation in 96 wells. Type I collagen, fibronectin, and α-smooth muscle actin were measured by ELISA after sonication of lung fibroblasts cultured in 24-well plates.

Nonlavaged whole lungs were collected and inflated with 3.6% buffered formaldehyde (Sigma-Aldrich). After overnight fixation, lungs were embedded in paraffin. Five-micrometer-thick sections were stained with H&E or Masson’s Trichrome.

Treatment-related differences were evaluated using unpaired t tests or one-way ANOVA, followed by pairwise comparisons with the Student–Newman–Keuls test, as appropriate (GraphPad, San Diego, CA). Statistical significance was considered at p < 0.05.

We first explored in C57BL/6 mice whether the proinflammatory cytokine IL-17A was expressed during the short- and long-term lung responses to silica particles. At days 3, 30, and 60 after silica treatment, we found a significantly increased pulmonary expression of IL-17A mRNA (quantitative RT-PCR [qRT-PCR]) in lung homogenates from C57BL/6 mice compared with saline controls (Fig. 1A). IL-17A protein levels in BALF were also increased after silica treatment at days 3 and 60 (Fig. 1B, 1C). Interestingly, the amplitude of silica-induced IL-17A upregulation was related to the susceptibility of mouse strains, because silica-resistant BALB/c mice showed the lowest IL-17A levels in BALF at days 3 and 60, whereas the highest levels were noted in highly susceptible DBA/2 mice (Fig. 1B, 1C). In C57BL/6 mice, we further detailed the pulmonary expression of IL-17A by assessing early lung responses to silica (6–72 h). Like for master proinflammatory cytokines (TNF-α [Fig. 2E] and IL-1β [Fig. 2F]) and chemokines (KC [Fig. 2H] and MIP-2 [data not shown]), the pulmonary IL-17A upregulation was already found 6 h after silica treatment (mRNA [Fig. 2A] and protein [Fig. 2D]); it preceded silica-induced tissue injury (Fig. 2I, BALF proteins), as well as leukocyte lung influx (Fig. 2J–L). The pulmonary expression of the IL-17A–related cytokines IL-23p19 and -6 (Fig. 2C, 2G) but not TGF-β (data not shown) coincided with IL-17A production (Fig. 2D). In contrast to IL-17A, lung expression of IL-17F was only transiently increased in silica-treated mice and lasted only 6 h (Fig. 2B). Altogether, these results indicated that silica induced a rapid and constant overexpression of IL-17A concomitantly with proinflammatory cytokines and chemokines and suggested a potentially important role for this cytokine during acute and chronic lung responses to silica.

To determine which lung cell type produced IL-17A after silica treatment, specific staining of immune cell markers and intracellular staining for IL-17A were performed at day 3, when IL-17A production and cellular inflammation were concomitant, after silica exposure (Fig. 2). FACS analysis showed that silica exposure enhanced the number of IL-17A⁺ lung cells 4-fold (Fig. 3A, 3E). More than 70% of these IL-17A⁺ lung cells were CD5⁺ T lymphocytes (data not shown). As shown in Fig. 3B–D and 3F–H, γδ⁺ and, to a lesser extent, CD4⁺, but not NK⁺ and CD8⁺, T lymphocytes (data not shown) produced IL-17A. To confirm these observations, we examined the effect of silica particles on IL-17A expression in athymic (nude) mice. Nude mice expressed markedly less IL-17A than the corresponding NMRI wild-type mice (Fig. 3I). In C57BL/6 mice, deficiency in γδ T cells or Ab depletion of CD4 T cells abrogated the pulmonary IL-17A expression induced by silica at day 3 (Fig. 3J, 3K). Finally, MACS separation of the different leukocyte populations present in the lung tissue at day 3 after silica treatment revealed that IL-17A mRNA expression was essentially found in γδ⁺ and CD4⁺ T cells and not in CD8⁺ T cells, macrophages and dendritic cells (CD11c⁺), or neutrophils (GR1⁺) (Fig. 3L). Purified lung fibroblasts or epithelial cells did not significantly express IL-17A (data not shown). Altogether, these data demonstrated that IL-17A–expressing cells that accumulated in the lung upon silica administration were essentially γδ⁺ T lymphocytes and CD4⁺ T cells.

We further examined by which mechanism IL-17A–producing T lymphocytes were recruited in the inflamed lungs. IL-23 is a cytokine that plays an important role in the induction of IL-17A expression, and we found that IL-23p19 and -23p40 in purified macrophages from silica-treated mice were increased 5-fold at day 3 (data not shown). It has also been shown that naive T cells differentiate to IL-17–producing T cells and expand in the presence of TGF-β and IL-6 (35, 36, 26). Silica-activated macrophages constitute a major source of these IL-17A–regulating cytokines (37). To determine whether macrophages contribute to IL-17A expression in the lungs, we examined the pulmonary expression of IL-17A in silica-treated mice depleted in macrophages through the administration of clodronate liposomes. IL-17A induction upon silica exposure was completely abrogated in macrophage-depleted mice (Fig. 4A). The abolished expression of IL-17A in IL-23p19–deficient mice (Fig. 4B) or after treatment with neutralizing anti–IL-23p40 Abs (data not shown) indicated a pivotal role for IL-23 in the expansion of IL-17A–producing T lymphocytes in this silicosis model. Blocking Abs against TGF-β and/or IL-6 failed to reduce the upregulation of IL-17A expression after silica (Fig. 4C). In conclusion, the accumulation of pulmonary IL-17–producing T cells depends on IL-23 expression by silica-activated macrophages.

Our first set of data suggested that recruited IL-17A–producing T lymphocytes might contribute to lung inflammation and fibrosis. To further address this hypothesis, we examined lung responses to silica in mice deficient for IL-17R or mice injected with neutralizing anti–IL-17A Abs at day 3, when lung inflammation and inflammatory mediators were present. Early lung inflammation was significantly reduced in silica-treated IL-17R–deficient mice compared with wild-type animals, as measured at day 3 by neutrophil counts (Fig. 5A) and total protein levels in BALF (Fig. 5B). In addition, reduced MPO activity was found in deficient mice, reflecting reduced neutrophil tissue activation and accumulation (e.g., IL-17R^+/+ mice, saline [OD] = 0.25 ± 0.03 and silica = 0.53 ± 0.04; IL-17R^−/− mice, saline = 0.21 ± 0.02 and silica = 0.35 ± 0.06). Administration of anti–mIL-17A Abs also controlled alveolitis (neutrophil counts [Fig. 5C]; protein levels [Fig. 5D]; and MPO [data not shown]). Limited cellular exudates filling the alveolar space and tissue were detected in IL-17R–deficient mice (Fig. 5I–L) compared with wild-type animals. Moreover, in the absence of IL-17R or after IL-17A Ab neutralization, the levels of proinflammatory cytokines IL-1β (Fig. 5E) and IL-6 (Fig. 5F), but not TNF-α (Fig. 5H), and the levels of chemokines KC (Fig. 5G) and MIP-2 (data not shown) were reduced. Assessment of lung inflammatory markers (tissue damage, cytokines, and chemokines) at day 1 in silica-treated IL-17R^−/− mice also revealed a significant reduction in acute alveolitis (data not shown). Finally, pharyngeal administration of recombinant mouse IL-17A (50–1000 ng/50 μl) significantly induced neutrophil influx in silica-treated BALB/c mice (data not shown). Collectively, these data suggested that IL-17A and IL-17A–producing T cells play a significant role in the establishment of the early alveolitis induced by silica to stimulate the production of proinflammatory chemokines and cytokines.

In this experimental model of silicosis, lung injury and neutrophil accumulation are not limited to day 3, but chronic and persistent inflammation develop with lung fibrosis (24). Therefore, we tested whether IL-17A plays a role in the chronic inflammatory lung responses induced by silica. At 60 d after silica, IL-17A did not appear to play a role in the long-term inflammatory process, because the responses were not attenuated in IL-17R^−/− mice or upon Ab IL-17A blockade (data not shown).

IL-22 was identified as a new Th17 cytokine (19, 12, 38). Thus, we questioned whether this cytokine could also be induced after silica treatment and whether it could represent an important factor in the establishment of alveolitis. IL-22 expression in the lung estimated by qRT-PCR was increased at early time-points after silica (day 3, Fig. 6A). IL-22 expression was largely T cell dependent because nude mice expressed significantly less IL-22 mRNA than did NMRI mice (Fig. 6B). In contrast to IL-17A, IL-22 expression was not critical for silica-induced alveolitis. Indeed, silica-treated IL-22–deficient mice developed similar early (Fig. 6C, 6D) and late (data not shown) alveolitis compared with wild-type animals.

Because it was reported that fibroblasts express receptors for Th17 cytokines and could be activated by IL-17A and -22 (39, 40), we assessed the direct effects of recombinant mouse (rm) IL-17A or -22 on the proliferation of lung fibroblasts purified from naive and silica-treated C57BL/6 mice (day 60) to determine the potential involvement of these Th17 cytokines in lung fibrosis. rmTGF-β1 (used as positive control) induced a strong proliferation of naive and silicotic fibroblasts (Fig. 7A). rmIL-17A, but not rmIL-22, significantly stimulated lung fibroblast proliferation (Fig. 7B, 7C) only in cultures obtained from silica-treated mice. Contrary to rmTGF-β1, Th17 cytokines did not induce collagen or fibronectin expression or myofibroblast differentiation (data not shown).

Because IL-17A was able to stimulate fibroblast proliferation in vitro, we next examined whether IL-17A–producing T lymphocytes might orchestrate the fibrotic process induced by silica in mice. The extent of silica-induced lung fibrosis, estimated by measuring the pulmonary levels of OH-proline (Fig. 8A), soluble collagen, type I collagen, and fibronectin, as well as by histology (data not shown), at day 60 appeared similar in IL-17R^−/− and IL-17R^+/+ mice. In addition, neutralizing mouse anti–IL-17A Abs were not able to modulate silica-induced lung fibrosis (e.g., OH-proline levels; irrelevant Abs, saline = 113.2 ± 3.1 μg/lung and silica = 193.7 ± 11.9 μg/lung; anti–IL-17 Ab, saline = 111.9 ± 4.1 μg/lung and silica 187.6 ± 10.3 μg/lung). The TGF-β lung levels, which increased upon silica treatment, were not reduced in the absence of IL-17 signaling (Fig. 8B). Similar data were observed in IL-22–deficient mice (Fig. 8C). Collectively, our observations showed that IL-17A and -22 are dispensable for the fibrotic lung response induced by silica in mice.

In silicosis, lung neutrophils have an important role in the development of cellular and DNA injury (41, 42), and several studies proposed that radicals and enzymes produced by neutrophils lead to lung destruction and cancer (43, 44). Activated macrophages and epithelial cells release proinflammatory mediators that induce neutrophil differentiation from the hematopoietic progenitors, their recruitment, and activation inside the inflamed lung tissue (45). In this study, we demonstrated an additional critical role for IL-17–producing T lymphocytes in the early neutrophilic response in an experimental model of silicosis.

The Th17 cell was reported to be a major producer of IL-17A. However, new cellular sources have been proposed, such as macrophages, neutrophils, and CD8, NK, αβ, and γδ T cells (13–15). In this study, we demonstrated that among the IL-17A–producing cells recruited upon silica treatment in the lung, only γδ⁺ and CD4⁺ T cells significantly expressed IL-17A. Indeed, the demonstration that nude and γδ T cell-deficient mice were not able to produce IL-17A in response to silica strongly supports the fact that γδ⁺ T cells represent the major cellular source of IL-17A. FACS analyses and purification procedures associated with the fact that anti-CD4 Ab significantly reduced IL-17A expression revealed that CD4⁺ T cells represent an additional IL-17A⁺ lung cell. The complete abrogation of IL-17A upregulation in γδ⁺ T lymphocyte-deficient mice also suggest that this particular population is important for IL-17A⁺ CD4⁺ T cell functions. This observation is reminiscent of the study by Simonian et al. (46), showing a possible interplay between γδ⁺ and CD4⁺ T cells in a murine model of hypersensitivity of pneumonitis.

It is still not clear which mediators control IL-17–producing T lymphocytes in vivo under inflammatory conditions. Recently, different investigators demonstrated that TGF-β acting in the presence of proinflammatory cytokines, particularly IL-6, is sufficient to differentiate and expand Th17 cells (26, 35, 36). Interestingly, it has been known for decades that macrophages activated by silica produce significant amounts of both cytokines, suggesting that these mediators may regulate IL-17–producing T cell differentiation and activation after silica exposure (37, 45). Our experimental data suggest that other mediators are important in the lungs. Neutralization of TGF-β and IL-6 activity was not associated with a reduction of lung IL-17A expression. Rather than expanding or stabilizing pre-existing IL-17–producing T lymphocytes, IL-23 has an earlier role during their differentiation in vivo, because IL-23R is quickly upregulated in response to proinflammatory cytokines (10). We demonstrated in this study that a similar activation occurred in silica-treated mice and that the accumulation of IL-17A–producing cells in the lung largely depended on IL-23 expression. Our finding is in accordance with recent studies establishing the pivotal role of IL-23 in Th17 cell influx into lung tissue in several other models of inflammation (14, 47–49).

Previous experimental studies using athymic mice provided important data supporting the paradigm that T lymphocytes are key immune cells in the establishment of neutrophilic responses induced by silica (50, 51). Our findings demonstrate a critical role for IL-17A–producing T cells in silica-induced neutrophilic inflammation and confer to this particular subpopulation a new pivotal role in lung neutrophilia during acute silicosis. It is well accepted that the proinflammatory functions of IL-17 and -22 are due to their capacities to stimulate neutrophilic factors released by epithelial cells. KC and MIP-2, two IL-8–related cytokines and powerful neutrophil chemoattractants, are produced by alveolar epithelial cells after silica exposure (52, 53), and neutralization experiments showed that both chemokines play an important role in the accumulation of lung neutrophils in silica-treated rats (54). The reduced production of KC and MIP-2 in the lungs of silica-treated IL-17R–deficient mice newly supports that, in addition to silica, IL-17A–producing T lymphocytes exacerbate proinflammatory functions of epithelial cells during silicosis. Silica-stimulated macrophages produce large amounts of proinflammatory cytokines, such as IL-6 and -1β and TNF-α, which are essential for the neutrophilic infiltration of the lung (6, 55). Interestingly, our data also suggest that, in addition to its proinflammatory activity on epithelial cells, IL-17A may amplify the production of the proinflammatory cytokines IL-1β and -6 by stimulated macrophages. Based on these observations, we propose that macrophages activated by silica produce powerful proinflammatory cytokines and induce an accumulation of IL-17A–producing T cells. In turn, these cells trigger the production of chemokines and cytokines by stimulated macrophages and epithelial cells that contribute to early lung neutrophil influx and injury.

We also showed that IL-17–producing T cells are not critical for chronic inflammation, because IL-17 deficiency did not affect inflammation intensity at later time-points (day 60). However, IL-17–related immune responses may participate in the development of many other chronic lung diseases also characterized by neutrophil-dominated accumulation, such as chronic obstructive pulmonary disease, cystic fibrosis, and certain forms of asthma, in humans and animals (56). Thus, it remains to be explained why chronic lung inflammation induced by silica is not directed by IL-17–producing T cells as it is during the earlier response. We demonstrated previously that long-term responses to silica in mice are strongly associated with the upregulation of suppressive cytokines, such as IL-10, which control, at least in part, the establishment of chronic inflammation (24, 57). Thus, we can speculate that this anti-inflammatory response can control the proinflammatory activities of IL-17A, as demonstrated in several recent reports (13, 58–60).

Pulmonary immunity appears essential in the establishment of numerous fibroproliferative lung disorders (61). However, the exact participation of the different Th cell subpopulations in the pathological process of lung fibrosis has not been clearly established (61). In experimental models of xenobiotic-induced lung fibrosis, contradictory results were generated by abrogating Th1 or Th2 effector cytokines. For instance, the absence of IFN-γ or IL-4 in mice was not associated with a unequivocal modulation of lung fibrosis (62–65). This may reflect the fact that another Th population, rather than Th1 and Th2, is involved in murine lung fibrosis. Interestingly, as for Th1 and Th2 cytokines, it was first suggested from in vitro and genome-wide searching approaches that the Th17 cytokines IL-17A and -22 could be involved in tissue-repair and wound-healing responses (11). Fibroblasts express the receptors for these cytokines, and IL-17A and -22 directly modulate fibroblast inflammatory functions by increasing IL-6 and -8, G-CSF, and matrix metalloproteinase production (39, 40, 66–70). In vitro studies also showed that rIL-17A did or did not upregulate the proliferation and collagen expression of cardiac or skin fibroblasts, respectively (39, 66). Few in vivo data have documented the hypothetical profibrotic functions of Th17-related cytokines in abnormal fibroproliferative repair. Some evidence from clinical observations indicated that uncontrolled Th17 responses underlie the development of lung fibrosis. IL-17A levels in lymphocytes and serum were elevated in patients developing systemic sclerosis and fibrosis (39). Concomitant increases in TGF-β, types I and III collagen, and IL-17A expression were also noted in patients with chronic asthma, as assessed by immunocytochemistry in bronchial biopsies (71). Finally, elevated IL-23 serum levels were associated with the prevalence of pulmonary fibrosis in systemic sclerosis patients (72). We also observed a strong association between IL-17A production and the establishment of lung fibrosis. However, our data clearly indicated that this association is not causal, because lung fibrosis can occur in the absence of Th17 cytokines. rIL-17A slightly increased fibrosis-related parameters in pulmonary fibroblasts obtained from silica-treated mice. However, this effect was not observed in naive fibroblasts and was not commensurable with the dramatic effect of TGF-β, a well-characterized profibrotic cytokine. However, it remains to be explained why, contrary to Ag-induced hypersensitivity pneumonitis and lung fibrosis, experimental lung fibrosis induced by silica is not directed by IL-17–producing T cells in mice (73, 74, 46). Very recent studies provide evidence that IL-17 and TGF-β are not obligatory related cytokines, as proposed before (49). It is well known that TGF-β is pivotal for the establishment of lung fibrosis in humans and experimental animals. We demonstrated in this study that TGF-β upregulation in experimental silicosis is independent of IL-17 expression and signaling, which may help to explain the unmodified lung fibrosis in IL-17R–deficient mice.

Altogether, our findings suggest that Th17 cells producing IL-17A are required for acute alveolitis during experimental silicosis. However, IL-17A and -22 are dispensable for chronic inflammation and are not essential modulators of fibroblast functions, uncontrolled tissue repair, and lung fibrosis induced by silica particles.

Disclosures The authors have no financial conflicts of interest.