NLRP3/Caspase-1–Independent IL-1β Production Mediates Diesel Exhaust Particle-Induced Pulmonary Inflammation

Inhalation of particulate matter pollutants is associated with several adverse health effects. Diesel exhaust particles (DEP) are a major component of this ambient particulate matter. Diverse studies in humans and mice demonstrated that DEP inhalation induces an inflammatory response in the lung (1–3). We previously showed in mice that DEP inhalation induces neutrophil and inflammatory monocyte recruitment to the bronchoalveolar lavage fluid (BALF) and lung tissue, and that DEP increases proinflammatory cytokine and chemokine expression in the lung. In addition, we demonstrated that DEP exposure results in increased dendritic cell (DC) accumulation and maturation in the lung. Exposure to DEP also increased DC migration to and enhanced the T cell response in the mediastinal lymph node (4). However, the molecular mechanisms leading to this DEP-induced pulmonary inflammation are largely unclear (1).

IL-1β is a proinflammatory cytokine that critically mediates diverse sterile inflammatory responses (5). Inhalation of particulate matter pollutants, including silica, asbestos, and cigarette smoke, induces IL-1β production in the lung (6–8). Moreover, upon exposure to these irritants, IL-1RI knockout (KO) mice showed severely impaired pulmonary inflammatory responses (6, 8). It was shown in vitro that DEP stimulate the IL-1β production from monocytes and DC (9, 10), macrophages (11), and epithelial cells (12). However, the functional contribution of IL-1β/IL-1RI signaling to DEP-induced lung inflammation in vivo was unstudied to date.

IL-1β is produced as an inactive cytoplasmatic precursor, pro–IL-1β, and requires proteolytic cleavage to become bioactive. In general, active caspase-1 (also known as the IL-1β–converting enzyme) cleaves IL-1β (13). Procaspase-1, in turn, is activated within cytoplasmatic multiprotein complexes, the inflammasomes. The NLRP3 inflammasome (NOD-like receptor, pryin domain containing 3) is unique in that it can be activated in response to endogenous danger signals and environmental irritants, including silica and asbestos particulates (14). Interestingly, NLRP3 KO mice that were exposed to silica and asbestos showed reduced BALF IL-1β levels and inflammation when compared with wild-type (WT) mice (7, 15). Whether NLRP3 and caspase-1 also mediate the IL-1β production and inflammatory response toward other particulate pollutants, such as DEP, was currently unclear.

In this study, we examined the role of the NLRP3/caspase-1/IL-1β pathway in the DEP-induced pulmonary inflammation.

Female IL-1RI KO mice and C57BL/6 WT mice were obtained from The Jackson Laboratory (Bar Harbor, ME) (16). Female caspase-1 KO mice and corresponding C57BL/6 WT mice were obtained from the Flanders Institute for Biotechnology/Ghent University (Ghent, Belgium) (17). Female NALP3 KO mice and corresponding C57BL/6 WT mice were obtained from Dr. Tschopp (Department of Biochemistry, University of Lausanne, Epalinges, Switzerland) and bred in our animalarium (18). All in vivo manipulations were approved by the Animal Ethical Committee of the Faculty of Medicine and Health Sciences of Ghent University.

DEP (SRM 2975) were obtained from the National Institute for Standards and Technology (Gaithersburg, MD). DEP were suspended in saline (B. Braun Melsungen, Melsungen, Germany) containing 0.05% Tween 80 (Invitrogen, Paisley, U.K.) to a concentration of 2 mg/ml. Mice were anesthetized with a ketamine/xylazine i.p. injection (70 mg/kg ketamine, Ketamine 1000 CEVA, Ceva Sante Animale, Libourne, France; 7 mg/kg xylazine, Rompun 2%, Bayer, Leverkusen, Germany) prior to instillation. The anesthetized mice were intratracheally instilled with 50 μl saline or DEP (=100 μg) solution at days 1, 4, and 7. On day 9, the animals were sacrificed by a lethal dose of i.p. pentobarbital (Ceva Sante Animale) (4).

At day 1, WT mice were intratracheally instilled with saline or 1 mg anakinra (Kineret; Biovitrum, Stockholm, Sweden), at 10 min before DEP instillation. For the IL-1β neutralization experiment, mice were intratracheally instilled with 50 μg anti–IL-1β (B122) or Armenian hamster IgG isotype control (HTK888) (both from BioLegend, San Diego, CA), at 10 min before DEP instillation. The animals were sacrificed on day 3 (4). We confirmed that this adapted protocol induced an inflammatory response comparable to the 9-d protocol.

A tracheal cannula was inserted, and BALF was recuperated by instillation of 3 × 300 μl HBSS without Ca²⁺ or Mg²⁺ (BioWhittaker, Lonza, Basel, Switzerland) supplemented with 1% BSA (for cytokine and chemokine measurements; Sigma-Aldrich, St. Louis, MO) and 6 × 500 μl HBSS without Ca²⁺ or Mg²⁺ supplemented with 0.6 mM sodium EDTA (Sigma-Aldrich). The lavage fractions were pooled. Total cell counts were performed using a Bürcker chamber.

Pulmonary circulation was rinsed with saline supplemented with EDTA to remove the intravascular pool of cells. Lungs were minced and incubated in digestion medium (RPMI 1640 supplemented with 5% FCS, 2 mM l-glutamine, 0.05 mM 2-ME [all Life Technologies, Invitrogen], 100 U/ml penicillin, 100 μg/ml streptomycin [Sigma-Aldrich], 1 mg/ml collagenase type 2 [Worthington Biochemical, Lakewood, NY], and 0.02 mg/ml DNase I [grade II from bovine pancreas; Boehringer Ingelheim, Ingelheim, Germany]) for 45 min at 37°C and 5% CO₂. RBCs were lysed using ammonium chloride (ACK) buffer. Cell counting was performed with a Z2 Coulter counter (Beckman Coulter, Fullerton, CA).

All staining procedures were performed in PBS without Ca²⁺ or Mg²⁺ (BioWhittaker, Lonza) containing 5 mM EDTA and 1% BSA. To minimize nonspecific bindings, BALF cells and lung single-cell suspensions were preincubated with anti-CD16/CD32 (clone 2.4G2). Cells were labeled with various combinations of mAbs. Neutrophils were characterized as CD11b⁺ (allophycocyanin-conjugated, clone M1/70), Ly6C⁺ (FITC-conjugated, clone AL-21), Ly6G⁺ (PE-conjugated, clone 1A8), F4/80⁻ (biotin-conjugated, clone BM8) cells. Inflammatory monocytes were CD11b⁺ (allophycocyanin-conjugated, clone M1/70), Ly6C⁺ (FITC-conjugated, clone AL-21), Ly6G⁻ (PE-conjugated, clone 1A8), F4/80⁺ (biotin-conjugated, clone BM8) cells. DC were characterized as CD11c⁺ (allophycocyanin-conjugated, clone HL3), low autofluorescent, MHCII⁺ (biotin-conjugated, clone 2G9) cells. Except F4/80, which was obtained from eBioscience (San Diego, CA), all Abs were obtained from BD Pharmingen (San Diego, CA). Data acquisition was performed on a FACSCalibur flow cytometer running CellQuest software (BD Biosciences, San Jose, CA). FlowJo software (Tree Star, Ashland, OR) was used for data analysis.

RNA was extracted from BALF cells using the miRNeasy mini kit (Qiagen, Hilden, Germany). Expression of neutrophil elastase, proteinase-3, and cathepsin G mRNA, relative to hypoxanthine guanine phosphoribosyltranferase mRNA, was analyzed using TaqMan gene expression assays (Applied Biosystems, Foster City, CA). RT-PCR was performed starting from 100 ng total RNA, using a LightCycler 480 Real-Time PCR System (Roche, Basel, Switzerland). Reverse transcription was performed by murine leukemia virus reverse transcriptase (Applied Biosystems) at 48°C for 30 min. Amplification conditions were as follows: 10 min at 95°C, and 45 cycles of 10 s at 95°C and 15 s at 60°C.

BALF, IL-1β, keratinocyte-derived chemokine (KC), MCP-1, and MIP-3α levels were measured using Quantikine ELISA kits (R&D Systems, Abingdon, U.K.).

Statistical analysis was performed with SPSS, version 17.0 (SPSS, Chicago, IL). Groups were compared using nonparametric tests (Kruskal-Wallis and Mann-Whitney U), according to standard statistical criteria. Reported values were expressed as mean ± SEM. The p values <0.05 were regarded as significant.

IL-1RI signaling is important in several types of sterile immune responses (5). To examine whether IL-1RI is also crucial in the DEP-induced inflammation, we exposed WT and IL-1RI KO mice to saline or 100 μg DEP. DEP exposure increased neutrophil, inflammatory monocyte, and DC numbers in BALF of WT mice, which is in accordance with our previous report (4). In IL-1RI KO mice, however, the DEP-induced neutrophil (Fig. 1A), inflammatory monocyte (Fig. 1B), and DC (Fig. 1C) accumulation was attenuated when compared with WT mice. DEP exposure also increased the neutrophil (Fig. 1D), inflammatory monocyte (Fig. 1E), and DC (Fig. 1F) counts in digested WT lungs, a response that was severely impaired or absent in IL-1RI KO mice. We also examined the levels of chemokines that attract these inflammatory cells. IL-1RI KO mice showed reduced KC (Fig. 1G), MCP-1 (Fig. 1H), and MIP-3α (Fig. 1I) concentrations in BALF compared with WT mice.

To further confirm the role of the IL-1RI in the DEP-induced inflammation, we instilled WT mice with a rIL-1RI antagonist (anakinra, Kineret) or saline, and subsequently exposed the mice to 100 μg DEP. Anakinra treatment strongly attenuated the DEP-induced neutrophil (Fig. 2A), inflammatory monocyte (Fig. 2B), and DC (Fig. 2C) accumulation in BALF. In agreement with this, concentrations of KC (Fig. 2D), MCP-1 (Fig. 2E), and MIP-3α (Fig. 2F) in BALF were reduced upon IL-R1 antagonism.

In vitro experiments have previously shown that DEP stimulation induces IL-1β (1). We first examined whether DEP induces IL-1β in our model. As shown in Fig. 3A, 100 μg DEP increased the IL-1β levels in BALF of WT mice. Next, to assess the functional contribution of IL-1β to the DEP-induced inflammation, we treated WT mice with an isotype control or anti–IL-1β mAb, and subsequently exposed the mice to 100 μg DEP. IL-1β neutralization reduced the DEP-induced neutrophil (Fig. 3B), inflammatory monocyte (Fig. 3C), and DC (Fig. 3D) accumulation in BALF.

Upon asbestos and silica exposure, it has been shown that NLRP3 activates caspase-1, which in turn matures IL-1β (7). By quantitative RT-PCR, we examined NLRP3 and caspase-1 mRNA expression upon DEP exposure. As shown in Fig. 4A and 4B, respectively, DEP increased NLRP3 mRNA and caspase-1 mRNA expression in BALF cells.

We next examined whether NLRP3 and caspase-1 are crucial for IL-1β production upon DEP exposure. For that purpose, NLRP3 KO, caspase-1 KO, and their corresponding WT mice were exposed to saline or 100 μg DEP. We observed that DEP-induced IL-1β production was similar between WT and NLRP3 KO mice (Fig. 5A). Also, in caspase-1 KO mice, the levels of IL-1β were comparable to WT mice (Fig. 5B). Because the (NLRP3) inflammasome/caspase-1 pathway also cleaves pro–IL-18 into mature IL-18, we assessed levels of this cytokine in our experiments. DEP exposure increased the IL-18 concentration in BALF of WT mice. In both NLRP3 KO and caspase-1 KO mice, however, DEP-induced IL-18 production was greatly impaired compared with WT mice (Fig. 5C, 5D, respectively).

We then further investigated the contribution of the NLRP3/caspase-1 pathway on the DEP-induced pulmonary inflammation. Neutrophil (Fig. 5E), inflammatory monocyte (Fig. 5F), and DC (Fig. 5G) counts in BALF upon DEP exposure were similar between WT and NLRP3 KO mice. Caspase-1 KO mice also showed comparable lung inflammation to WT controls (Fig. 5H–J).

The above-described observations suggest a NLRP3/caspase-1–independent pathway for IL-1β production upon DEP exposure. In neutrophil-mediated inflammatory responses, it was demonstrated that neutrophil-derived serine proteases can cleave pro–IL-1β (13, 19). We therefore investigated the serine protease expression upon DEP using quantitative RT-PCR. Exposure to DEP largely increased the neutrophil elastase (Fig. 6A), proteinase-3 (Fig. 6B), and cathepsin G (Fig. 6C) mRNA expression in BALF cells.

Environmental pollutants like DEP cause inflammatory reactions in the lung; however, their mechanisms of action were largely unclear (1). In this study, we demonstrate that the DEP-induced pulmonary inflammation critically depends on IL-1β/IL-1RI signaling. Furthermore, we examined the mechanism of the IL-1β production upon DEP exposure. We show that the NLRP3 inflammasome/caspase-1 pathway is not necessary for the DEP-induced IL-1β production, suggesting that alternative pathways are implicated upon DEP exposure.

IL-1 signaling through IL-1RI leads to the activation of the transcription factors NF-κB and AP-1, which then finally leads to expression of vascular adhesion molecules, the induction of chemokines like KC and MIP-3α, and the recruitment of inflammatory cells to tissues (13). Previously, we showed that DEP exposure induces a pulmonary inflammatory response, characterized by neutrophil, inflammatory monocyte, and DC accumulation in the airways and lung, and increased KC, MCP-1, and MIP-3α chemokine levels in BALF (4). In this study, we demonstrate that this DEP-induced inflammation was severely reduced in IL-1RI KO mice. Using a human rIL-1R antagonist (anakinra), we further confirmed the importance of IL-1RI signaling in the DEP-induced inflammation. IL-1RI signaling is also crucial in several other sterile immune responses (5). For example, upon smoke (8, 20), ozone (21), and silica exposure (6), the induced lung inflammation was IL-1RI dependent. The observation that these sterile stimuli all signal through the IL-1RI could suggest that pollutants stimulate lung inflammatory responses through a common IL-1RI–dependent mechanism (5). However, DEP inhalation can also induce chemokines, like KC, directly from the epithelium (22). Probably, this explains the limited, but increased, pulmonary inflammation in DEP-exposed IL-1RI KO mice when compared with saline-exposed IL-1RI KO mice.

We found increased IL-1β levels in BALF upon DEP exposure. This is in line with the finding that particulate matter instillation is associated with increased IL-1β in BALF of healthy persons (23). In several in vitro experiments, various cells such as inflammatory monocytes and DC (9, 10), macrophages (11), and epithelial cells (12) were shown to produce IL-1β upon DEP exposure. Furthermore, neutrophils, which are recruited upon DEP inhalation, can also produce IL-1β (13).

Both IL-1α and IL-1β isoforms signal through the IL-1RI (13). We examined the role of IL-1β in the DEP-induced inflammation by treating WT mice with an IL-1β–neutralizing mAb and demonstrated that this impaired the inflammatory response. However, these experiments do not exclude a role of IL-1α in the DEP-induced inflammation, because we also observed increased IL-1α levels upon DEP exposure (data not shown). In other sterile inflammatory responses, contributions for both IL-1α or IL-1β were demonstrated (5). A recent paper studying the relative contribution of these isoforms in the sterile inflammation induced by death cells showed that both IL-1α and IL-1β were critically involved (24).

In general, it is proposed that caspase-1 cleaves pro–IL-1β into mature IL-1β. Caspase-1, in turn, requires activation within multiprotein complexes, the inflammasomes (13, 25). Whereas most inflammasomes are activated in response to pathogen-associated molecular patterns, the NLRP3 inflammasome is unique in that it is also activated in response to damage-associated molecular patterns and environmental irritants (14, 25). The airborne pollutants silica and asbestos, for example, were shown to activate NLRP3 in vitro (7). In addition, NLRP3 KO mice, which were exposed to these pollutants, showed reduced BALF IL-1β levels and inflammation when compared with WT mice (7, 15). Therefore, we investigated whether the DEP-induced IL-1β production and inflammation would also be NLRP3 dependent. We observed no functional involvement for NLRP3 in the DEP-induced response, despite increased NLRP3 mRNA expression upon DEP instillation, which is probably explained by increased inflammatory cell recruitment. Besides NLRP3, other inflammasomes can activate caspase-1 or diverse inflammasomes could interact to activate caspase-1 (25). For example, during Listeria monocytogenes infection, it was shown that several inflammasomes activate the caspase-1 (26). Using caspase-1 KO mice, we therefore further examined the contribution of the inflammasome/caspase-1 pathway to the DEP-induced IL-1β levels and inflammation. Again, we found no difference between the caspase-1 KO and WT mice. Taken together, these results suggest that the (NLRP3) inflammasome/caspase-1 pathway is redundant in the DEP-induced IL-1β production and inflammation. This suggests that alternative mechanisms might be involved in IL-1β maturation upon DEP exposure.

Inflammasome/caspase-1–independent IL-1β maturation was previously reported in some murine models (13, 19). Upon infection with Mycobacterium tuberculosis, IL-1β was shown to be produced in a caspase-1– and ASC (component of most inflammasomes)-independent manner (27); and also in a turpentine model, the sterile inflammation was impaired in IL-1β KO mice, but not in caspase-1 KO mice (28). Sterile inflammatory responses are often associated with massive neutrophil recruitment, and it is demonstrated that neutrophil-derived serine proteases (such as neutrophil elastase, proteinase-3, and cathepsin G) also can cleave pro–IL-1β into bioactive IL-1β (29, 30). For example, in a mouse model of acute arthritis, which is characterized by neutrophil recruitment, proteinase-3 mediates the caspase-1–independent IL-1β processing (31, 32). Upon LPS exposure, it is demonstrated that administration of a cell-permeable protease inhibitor prevented the IL-1β processing by neutrophils, whereas an extracellular inhibitor did not affect the IL-1β production (30). This suggests that neutrophil serine proteases can produce IL-1β intracellularly, in a mechanism paralleling the caspase-1–mediated IL-1β processing. Because exposure to DEP is associated with a massive accumulation of neutrophils and increased neutrophil serine protease expression, one could speculate that neutrophil serine proteases cleave the DEP-induced IL-1β. However, neutrophil depletion prior to DEP exposure resulted in compensatory mononuclear cell recruitment and IL-1β production (S. Provoost, T. Maes, L. Boon, G.F. Joos, and K.G. Tournoy, unpublished results), demonstrating that IL-1β production is complex and requires further investigation. In this respect, it should also be mentioned that besides serine proteases, other enzymes like chymase, chymotrypsin, granzyme A, and matrix metalloproteinase not restricted to the neutrophil can cleave IL-1β in vitro (33).

Besides IL-1β maturation, the NLRP3/caspase-1 pathway can also cleave pro–IL-18 into bioactive IL-18 (13). In our model, we found that the IL-18 concentrations in BALF were reduced in DEP-exposed NLRP3 KO and caspase-1 KO mice. However, because the pulmonary inflammation upon DEP was similar between these KO and WT mice, this suggests that this proinflammatory cytokine is not critical in the DEP-induced immune response. Interestingly, whereas it is in general thought that IL-1β and IL-18 mature through the same pathway, our findings suggest a differential regulation (inflammasome independent versus dependent, respectively) of these cytokines upon DEP exposure.

In conclusion, these data provide insight into the mechanisms that mediate the inflammatory responses upon diesel inhalation. We show that the IL-1β/IL-1RI signaling is critical in the pulmonary inflammation induced by DEP exposure. Furthermore, we demonstrate that the DEP-induced IL-1β production does not critically depend on the classical NLRP3 inflammasome/caspase-1 pathway. This suggests that other proteases could cleave the DEP-induced IL-1β.

We thank Prof. Tschopp (Department of Biochemistry, University of Lausanne, Epalinges, Switzerland) for providing us the NLRP3 KO mice. We also thank Eliane Castrique, Christelle Snauwaert, Evelyn Spruyt, Greet Barbier, Indra De Borle, Ann Neesen, Katleen De Saedeleer, Anouck Goethals, Marie-Rose Mouton, and Philippe De Gryze for technical assistance.

The authors have no financial conflicts of interest.