Inflammasome-Independent and Atypical Processing of IL-1β Contributes to Acid Aspiration–Induced Acute Lung Injury

Gastric aspiration pneumonitis is defined as an acute lung injury (ALI) following the inhalation of gastric contents and is associated with significant morbidity and mortality (1, 2). Aspiration pneumonitis commonly occurs as a complication of general anesthesia and in patients with altered levels of consciousness due to trauma, cerebral vascular ischemia, or metabolic encephalopathies. The inhalation of highly acidic gastric fluid and/or particulate food matter promotes chemical pneumonitis and the development of ALI, which can frequently be complicated by subsequent bacterial pneumonia. Previous studies have demonstrated that aspirated acid induces dysfunction of alveolar epithelial fluid transport, followed by alveolar epithelial cell injury and neutrophil infiltration in the lung (3, 4). However, the mechanisms that underline acid aspiration–induced ALI are unclear, because acid itself is rapidly neutralized by proteins and bicarbonate systems (4). Recent investigations have shown that inflammation plays a pivotal role in the progression of acid aspiration–induced ALI (1, 2). Because gastric fluid is highly acidic, it is considered that gastric aspiration triggers sterile inflammation and injury in the lung.

Recent evidence indicates that sterile inflammation is mediated through the nucleotide-binding oligomerization domain-like receptor (NLR) family pyrin domain containing 3 (NLRP3) inflammasome, an intracellular, large, multiple-protein complex that regulates the processing of a potent proinflammatory cytokine IL-1β (5, 6). NLRP3 inflammasome contains NLRP3, an adaptor molecule apoptosis-associated, speck-like protein containing a caspase recruitment domain (ASC), and cysteine protease caspase-1, which induces caspase-1 activation. Because caspase-1 is known to be an IL-1β–converting enzyme (7), its activation processes pro–IL-1β into its mature form and induces IL-1β production, thereby leading to an inflammatory response and tissue injury. Indeed, we recently demonstrated that NLRP3 inflammasome was involved in the pathogenesis of sterile inflammation-related diseases (8–12). In terms of acidic stress, a mild acidic environment (pH 6–6.5) frequently occurs at sites of inflammation and ischemia (13, 14). Several studies have suggested that an inflammatory response is influenced under mild acidic conditions. However, the role of NLRP3 inflammasome under such mild acidic conditions remains controversial. For instance, Rajamaki et al. (15) recently demonstrated that mild acidic stress triggered NLRP3 inflammasome activation and IL-1β production through K⁺ efflux in human macrophages. In contrast, Takenouchi et al. (16) found that mild acidic conditions modified P2X7 receptor–dependent IL-1β production in cultured murine microglial cells, independent of caspase-1. Similarly, Edye et al. (17) reported that mild acidic conditions promoted danger signal-induced IL-1β production in a caspase-1–independent manner in human monocytic THP-1 cells and murine glial cells. Intriguingly, the latter two reports suggested that acidic stress could alter pro–IL-1β processing site(s) and produce the 20-kDa mature form (p20–IL-1β) rather than the 17-kDa, caspase-1–dependent form (p17–IL-1β). They further showed that the aspartic protease cathepsin D is responsible for p20–IL-1β processing. However, no information is currently available on the role of NLRP3 inflammasome and IL-1β production under severe acidic conditions that simulate exposure to gastric fluid (pH < 2.5). In the current study, we used mice deficient in NLRP3 and IL-1β and unexpectedly found that the deficiency of IL-1β, but not NLRP3, resulted in fewer inflammatory responses and injury in the lung after severe acid aspiration, indicating that NLRP3 inflammasome-independent IL-1β contributes to the development of acidic stress–induced ALI. Furthermore, using mass spectrometry and processing-site mutants of pro–IL-1β, we identified a novel cleavage site of pro–IL-1β in response to severe acidic stress. Our findings demonstrate that IL-1β plays a novel role in aspiration pneumonitis and provide new insight into the mechanism underlying severe acidic stress–mediated inflammation.

All experiments in this study were approved by the Use and Care of Experimental Animals Committee of Jichi Medical University (permit number 17151) and conducted in accordance with Jichi Medical University guidelines. C57BL/6J (wild-type [WT]) mice were purchased from Japan SLC (Tokyo, Japan). NLRP3^−/− and IL-1β^−/− mice (C57BL/6J genetic background) were kindly provided by Dr. Vishva M. Dixit (Genentech, Southern San Francisco, CA) and Dr. Yoichiro Iwakura (Tokyo University of Science, Chiba, Japan), respectively (18, 19). Female mice aged 8−14 wk were used. The mice were intratracheally injected with 50 μl of 0.05 N hydrochloric acid (HCl) diluted by saline or vehicle (PBS) (4, 20, 21). We analyzed the lung samples at 24 h after acid aspiration. Mice were housed (four mice per cage; RAIR HD ventilated Micro-Isolator Animal Housing Systems, Lab Products; Seaford, DE) in an environment maintained at 23 ± 2°C with ad libitum access to food and water under a 12-h light and dark cycle with lights on from 8:00 to 20:00.

Bronchoalveolar lavage fluid (BALF) was obtained by cannulating the trachea with an 18-gauge catheter. After the whole lung was lavaged four times with 0.8 ml of PBS, the lavage fluid was centrifuged at 1000 rpm for 10 min at 4°C, and cell-free supernatants were stored at −30°C. The pellet was diluted in PBS, the cells were stained with trypan blue, and the number of live cells was counted using a hemocytometer. Differential cell analysis was performed by staining with Diff-Quik (Sysmax, Kobe, Japan) after a cytospin (800 rpm for 8 min at 22°C) and by flow cytometry.

RNA was extracted from the lungs perfused with PBS or cultured cells using ISOGEN (Nippon Gene, Toyama, Japan) according to the manufacturer’s instructions. Real-time RT-PCR analysis was performed using a Takara TP960 PCR Thermal Cycler Dice Real Time System (Takara Bio, Shiga, Japan) to detect mRNA. The primers (antisense and sense, respectively) were as follows: Il1b: 5′-TGAAGTTGACGGACCCCAAA-3′ and 5′-TGATGTGCTGCTGTGAGATT-3′, Ccl2: 5′-GGCTCAGCCAGATGCAGTTAAC-3′ and 5′-GCCTACTCATTGGGATCATCTTG-3′, Cxcl1: 5′-GCTGGGATTCACCTCAAGAA-3′ and 5′-TCTCCGTTACTTGGGGACAC-3′, 18S rRNA (Rn18s): 5′-GTAACCCGTTGAACCCCATT-3′ and 5′-CCATCCAATCGGTAGTAGCG-3′, CCL2: 5′-CAGCCAGATGCAATCAATGCC-3′ and 5′-TGGAATCCTGAACCCACTTCT-3′, CXCL1: 5′-GGAACAGAAGAGGAAAGAGAGAC-3′ and 5′-TAGGACAGTGTGCAGGTAGA-3′, and 18S rRNA (RNA18S5): 5′-GTAACCCGTTGAACCCCATT-3′ and 5′-CCATCCAATCGGTAGTAGCG-3′. Expression levels were quantified using a standard curve and were normalized to the content of 18S rRNA. Each normalized value was expressed as a ratio to the value for WT mice after intratracheal aspiration of PBS.

Cell lysates were prepared using radioimmunoprecipitation assay buffer (20 mM Tris, 2.5 mM EDTA, 1% Triton X, 10% glycerol, 1% deoxycholic acid, 0.1% SDS, 50 mM NaF, and 10 mM Na₄P₂O₇ × 10 H₂O), and subjected to SDS-PAGE. The proteins were electrophoretically transferred to a PVDF membrane. The membrane was blocked with 2% casein or 3% BSA for 1 h at room temperature and then incubated overnight at 4°C with primary Abs, followed by incubation for 1 h with secondary Abs conjugated to HRP. Immunoreactive bands were visualized using a Western BLoT HRP Chemiluminescent Substrate system (Takara Bio). The expression level of β-actin served as an internal control for protein loading. Primary Abs were used those against p38, phospho-p38, caspase-1 (Cell Signaling Technology, Boston, MA), IL-1β (R&D Systems, Minneapolis, MN), and anti–β-actin (Sigma-Aldrich, St. Louis, MO). HRP/goat anti-mouse IgG (Invitrogen, Carlsbad, CA) and HRP/goat anti-rabbit IgG (Zymax Grade, Zymed Laboratories, South San Francisco, CA) were used as secondary Abs. The results represent at least three independent experiments.

To extract proteins with a molecular mass between 10 and 30 kDa, the supernatants isolated from radioimmunoprecipitation assay lysate were ultrafiltrated using Amicon Ultra (MilliporeSigma, Darmstadt, Germany). After the ultrafiltrated supernatants were incubated with anti–IL-1β Ab (Santa Cruz Biotechnology, Dallas, TX) for 1 h at 4°C, Protein G Sepharose (GE Healthcare, Chicago, IL) was added, and the mixture was incubated overnight at 4°C. The proteins that bound to Protein G Sepharose were eluted by 0.1 M glycine HCl (pH 3.5).

IL-1β levels were measured using a mouse or human ELISA kit (R&D Systems) according to the manufacturer’s instructions. The bioactivity of IL-1β was analyzed using IL-1R antagonist (IL-1Ra; PeproTech, Rocky Hill, NJ).

Cells collected from BALF were analyzed using flow cytometry, as described previously (22). The cells were double labeled with the following Abs: allophycocyanin-conjugated anti-CD45 (eBioscience, San Diego, CA), FITC-conjugated anti-CD45R (eBioscience), PE-conjugated anti–Ly-6G (BD Biosciences, San Jose, CA), FITC-conjugated anti-CD11c (BD Bioscience), FITC-conjugated anti-F4/80 (eBioscience), and PE-conjugated anti-CD11b (eBioscience). The cells were examined by flow cytometry (FACS Verse; BD Biosciences); the analysis was performed using FlowJo software (version 10; Tree Star, San Carlos, CA). Isotype control Abs were used as negative controls to exclude nonspecific staining. Dead cells were identified by the 7-AAD (BD Biosciences) staining.

Lungs were fixed by the intratracheal injection of 1 ml of 10% formalin and embedded in paraffin. Lung tissue sections (5-μm thick) were stained with H&E. ALI was scored from four points: 1) alveolar congestion, 2) hemorrhage, 3) infiltration or aggregation of neutrophils in airspace or the vessel wall, and 4) thickness of the alveolar wall/hyaline membrane formation, as describing previously (23, 24). Each item was graded on a five-point scale: 0, minimal (little) damage; 1+, mild damage; 2+, moderate damage; 3+, severe damage; and 4+, maximal damage. Lung tissue sections were observed excluding the peribronchial area where inflammatory cells had infiltrated excessively.

Immunohistochemical analyses were performed to detect the pan-leukocyte marker CD45 and the oxidative stress marker 4-hydroxy-2-nonenal (4-HNE). In brief, deparaffinized sections were boiled in Target Retrieval Solution (Dako, Agilent Pathology Solutions, Santa Clara, CA), blocked with normal goat serum, and incubated overnight with an Ab against CD45 (BD Biosciences, Franklin Lakes, NJ). This was followed by incubation with Histofine Simple Stain Rat MAX PO (Nichirei Corporation, Tokyo, Japan). The immune complexes were detected using a DAB substrate kit (Vector Laboratories, Burlingame, CA). For 4-HNE immunostaining, the sections were blocked with mouse IgG blocking reagent (Mouse-On-Mouse Immunodetection Kit; Vector Laboratories) and incubated overnight with an Ab against 4-HNE (clone HNEJ-2, Japan Institute for the Control of Aging, Nikken SEIL, Shizuoka, Japan). This was followed by incubation with biotin-conjugated secondary Abs. The sections were treated with avidin–peroxidase (VECTASTAIN ABC Kit; Vector Laboratories). The reaction was developed using a DAB Substrate Kit (Vector Laboratories). The sections were counterstained with hematoxylin. No signals were detected when an irrelevant IgG (Vector Laboratories) was used instead of the primary Ab as a negative control. The images of the stained sections were obtained and analyzed using a microscope (FSX-100; Olympus, Tokyo, Japan). Five visual fields (0.2 mm²/field) were randomly selected to count positively stained cells.

Human monocytic THP-1 and human lung epithelial A549 cell lines were obtained from American Type Culture Collection (Manassas, VA). THP-1 cells were grown in RPMI-1640 (Sigma-Aldrich) supplemented with 10% FBS. A549 cells were grown in DMEM (glucose 1000 mg/L; Wako Pure Chemical Industries, Osaka, Japan) supplemented with 10% FBS. Murine peritoneal macrophages were isolated using the thioglycollate elicitation method and cultured in RPMI 1640 supplemented with 10% FBS. For acidic stress experiments, THP-1 cells were cultured in 24-well tissue culture plates and differentiated into macrophages with PMA (200 nM; Sigma-Aldrich) for 14 h. After being primed with or without LPS (100 ng/ml; Sigma-Aldrich) for 16 h, cells were treated with control (unbuffered balanced salt solution [UBSS]; 144 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl₂, and 0.8 mM MgCl₂) or acidic solution (UBSS [pH 1.75] adjusted by HCl), as previously described (25). Subsequently, culture media were washed with 100 μl UBSS and then treated with 100 μl UBSS or UBSS (pH 1.75). After 5 min, 400 μl of 0.1% BSA/RPMI 1640 was added, and the cells were cultured for 6 h. As a positive control of NLRP3 inflammasome activation, THP-1 macrophages were treated with adenosine 3′-phosphate (5 mM ATP; Sigma-Aldrich). The same protocol was used in primary murine macrophages.

For inhibitor experiments, cells were pretreated for 1 h with 4-(2-aminoethyl)-benzenesulfonyl fluoride hydrochloride (AEBSF; Roche Diagnostics, Dubai, United Arab Emirates), E64 (Sigma-Aldrich), cathepsin G inhibitor I (CGI; Cayman Chemical, Ann Arbor, MI), 20 μM CA-074 Me (Wako Pure Chemical Industries), Z-Tyr-Val-Ala-Asp-fluoromethyl ketone (20 μM) (Medical and Biological Laboratories, Nagoya, Japan), and Mito-TEMPO (80 μM; Enzo Life Sciences, Farmingdale, NY), and then treated with UBSS or UBSS (pH 1.75) as described above. For experiments with mild acidic stress, THP-1 macrophages were treated for 8 h with HEPES-buffered salt solution (HBSS, 145 mM NaCl, 2.5 mM KCl, 1 mM MgCl₂, 1.8 mM CaCl₂, 20 mM HEPES, and 10 mM glucose; [pH 7.4]), HBSS (pH 6.2) (adjusted by NaOH), and lactic acid (25 mM; MP Biomedicals, Santa Ana, CA), according to the previous reports (16, 26). Cells were pretreated for 15 min with the cathepsin D inhibitor pepstatin A (50 μM; Serva Electrophoresis, Heidelberg, Germany). All reagents were obtained from Sigma-Aldrich (St. Louis, MO) unless otherwise specified.

Mutated IL-1β_T107A, IL-1β_W108A, IL-1β_D109A, IL-1β_N110A, IL-1β_Y113D, IL-1β_V114E, and IL-1β_D116I were generated using the PrimeSTAR Mutagenesis Basal Kit (Takara Bio, Shiga, Japan) and subcloned into a CS-CA-MCS plasmid (kindly provided by Dr. H. Miyoshi, RIKEN BioResource Center, Ibaraki, Japan) (27) with a C-terminal FLAG-6 × His tag. To prepare the lentiviral vectors, HEK293T cells were cotransfected together with CS-CA-MCS, pLP1, pLP2, and pVSVG using PEI Max (Polysciences, Warrington, PA). Culture media containing the lentiviral vectors were collected and purified by ultracentrifugation. The lentiviral titer was measured using a Lentivirus qPCR Titer Kit (Applied Biological Materials, Richmond, BC, Canada). For the production of THP-1 cells expressing mutated IL-1β, cells were transduced in the presence of 8 μg/ml polybrene (Sigma-Aldrich). Transfected FLAG/IL-1β proteins were purified with Flag-M2 beads from cell lysates. Humanized Kusabira-Orange1 (hKO1) subcloned into CS-CA-MCS was used as a negative control.

From the silver-stained gel after electrophoresis, the band at the position of the mature form of IL-1β was excised and subjected to a standard in-gel trypsin digestion protocol (28). After the enzyme digestion solution was recovered, it was dried and redissolved in 0.1% formic acid solution. After centrifugation at 20,000 × g for 10 min at room temperature, the supernatant was recovered and subjected to liquid chromatography–tandem mass spectrometry (Q Exactive Plus, Thermo Fisher Scientific, Waltham, MA). Tandem mass spectrometry ion search was performed using Proteome Discoverer Software (Thermo Fisher Scientific). Extraction of the N-terminal peptide from the peptide fragments identified by a Mascot search (Matrix Science, Boston, MA) suggested that the N terminus is F104 or N110. In either case, the peptide bond on the N-terminal side of the residue cannot be cleaved by trypsin; therefore, both F104 and N110 are N-terminals cleaved by an endogenous enzyme in the cell. Furthermore, the theoretical molecular mass of cleaved IL-1β was calculated based on the predicted sequences using the “Compute pI/Mw” tool on the ExPASy server (http://web.expasy.org/compute_pi/).

NLRP3 and caspase-1 genes were edited by CRISPR/Cas9 system in THP-1 cells, which was performed as described previously (29). Briefly, human codon–optimized Cas9 and single-guide RNA (sgRNA)–expressing vectors (LentiCRISPRv2) were obtained from Addgene (Watertown, MA). The sgRNA-targeting NLRP3 and caspase-1 were designed with CRISPR direct (http://crispr.dbcls.jp). The target sequences of sgRNA (antisense and sense, respectively) were as follows: 5′-CACCG CAGGGTCAGCTTGCCGTAGG-3′ and 5′-AAACCCTACGGCAAGCTGACCTGC-3′ (GFP); 5′-CACCGGATCGCAGCGAAGATCCACA-3′ and 5′-AAACTGTGGATCTTCGCTGCGATCC-3′ (NLRP3); and 5′-CACCGAAGCTGTTTATCCGTTCCAT-3′ and 5′-AAACATGGAACGGATAAACAGCTTC-3′ (caspase-1). For lentiviral transduction, THP-1 cells were incubated with lentiviral vectors for 16 h in the presence of 8 μg/ml polybrene (Sigma-Aldrich). EGFP-, NLRP3-, and caspase-1–deleted cells were further selected by incubating them with 2 μg/ml puromycin (Sigma-Aldrich) for at least 3 d.

Data were analyzed using IBM SPSS Statistics Version 21 Software (IBM Japan, Tokyo, Japan) and expressed as the mean ± SEM. An unpaired t test was used to compare two groups following Levene test for homogeneity of variance. For comparisons between multiple groups, the significance of differences between group means was determined using one-way ANOVA combined with Tukey test or the Games–Howell test. A p value <0.05 was considered to be statistically significant.

To investigate the contributions of IL-1β and NLRP3 inflammasome in gastric aspiration–induced ALI, WT, IL-1β^−/−, and NLRP3^−/− mice were subjected to intratracheal aspiration of acid solution (50 μl of 0.05-N HCl) or vehicle, and analyzed 24 h after aspiration. H&E staining showed that acid aspiration caused alveolar congestion, hemorrhage, inflammatory cell infiltration, and alveolar wall thickening in WT mice, compared with vehicle aspiration (Fig. 1A). Quantitative analysis showed that all of these manifestations were significantly attenuated in IL-1β^−/− mice, but not in NLRP3^−/− mice (Fig. 1B). We next assessed IL-1β production in BALF of WT and NLRP3^−/− mice and found that acid aspiration significantly increased IL-1β levels in the BALF of WT mice. However, there was no significant difference in IL-1β production between WT and NLRP3^−/− mice (Fig. 1C). Real-time RT-PCR analysis showed the upregulation of Il1b mRNA in the lungs of WT mice after acid aspiration (Fig. 1D). These results indicate that IL-1β contributes to the development of acid aspiration–induced ALI independent of NLRP3 inflammasome.

To investigate inflammatory cell infiltration into the lungs, we performed immunohistochemical staining for the pan-leukocyte marker CD45. Acid aspiration clearly increased the number of CD45⁺ cells in the lungs of WT mice, and this increase was significantly prevented in IL-1β^−/− mice (Fig. 2A, 2B). We also assessed the infiltration of inflammatory cells, such as alveolar macrophages (F4/80⁺/CD11c⁺), activated macrophages (CD11b⁺/CD11c⁺), and neutrophils (CD45R⁻/Ly-6G⁺), in BALF. The number of total inflammatory cells, alveolar macrophages, activated macrophages, and neutrophils was significantly increased in WT mice after acid aspiration (Fig. 2C, 2D). The infiltration of total inflammatory cells and alveolar macrophages was significantly decreased in lungs of IL-1β^−/− mice. The infiltration of activated macrophages and neutrophils also tended to be decreased in IL-1β^−/− mice, but this difference was not statistically significant. Similar changes were observed in the expression of inflammatory chemokines, Ccl2 and Cxcl1 (data not shown). Because reactive oxygen species (ROS) contributes to the development of gastric aspiration–induced ALI (30), we performed an immunohistochemical analysis for the lipid peroxidation marker 4-HNE to assess ROS generation. 4-HNE⁺ cells were clearly visualized in the lung of WT mice after acid aspiration (Fig. 2E, 2F), and the number of these cells was markedly lower in the lungs of IL-1β^−/− mice. These results suggest that IL-1β plays a crucial role in the inflammatory response and ROS generation in acid aspiration–induced ALI.

To explore the mechanisms by which acid aspiration triggers IL-1β production in the lung, we examined the effects of acidic stress on IL-1β production in macrophages in vitro because alveolar resident macrophages are directly exposed to acidic stress when subjected to acid aspiration. Severe acidic stress, loaded by a brief exposure to acid solution (UBSS [pH 1.75]), markedly stimulated IL-1β production in PMA-differentiated THP-1 macrophages (Fig. 3A). There was no difference in acidic stress–induced IL-1β production between unprimed and LPS-primed cells; this result is reasonable because pro–IL-1β synthesis is induced by PMA alone (31). To determine the contribution of NLRP3 inflammasome in acidic stress–induced IL-1β production, the activation of caspase-1 was determined by western blotting. Although the well-known NLRP3 inflammasome activator ATP obviously caused the cleavage (p20) of caspase-1, acidic stress failed to induce its cleavage (Fig. 3B), indicating that NLRP3 inflammasome is not involved in severe acidic stress–induced IL-1β production. Additionally, although we detected weak bands of mature IL-1β in severe acidic stress–treated cells, we found marked bands of mature IL-1β in ATP-treated cells (Fig. 3C). To enhance the weak bands of the mature form, we performed immunoprecipitation using anti–IL-1β Ab and clearly showed that severe acidic stress mainly produced the 18-kDa (p18–IL-1β) mature form, which is different from 17 kDa (p17–IL-1β) processed by caspase-1. The large majority of mature IL-1β processed by ATP and acidic stress was p17–IL-1β and p18–IL-1β, respectively. Consistent with the findings in THP-1 macrophages, severe acidic stress also stimulated IL-1β production by primary macrophages from WT mice. This IL-1β production retained by macrophages from NLRP3^−/− or caspase-1^−/− mice (Fig. 3D). As expected, ATP-induced IL-1β production was almost completely abolished by the deficiency of NLRP3 and caspase-1. These findings suggest that severe acidic stress processes p18–IL-1β, which is distinct from caspase-1–processed p17–IL-1β in macrophages.

To confirm that acidic stress produced mature IL-1β independent of NLRP3 inflammasome, we tested the effects of inhibitors related to NLRP3 inflammasome: 130 mM KCl (potassium efflux inhibition), Z-YVAD (caspase-1 inhibitor), CA074-Me (cathepsin B inhibitor), and Mito-TEMPO (mitochondria-targeted antioxidant). Although none of these inhibitors inhibited severe acidic stress–induced IL-1β production, all except Mito-TEMPO significantly inhibited ATP-induced IL-1β production (Fig. 4A). Consistently, Western blotting showed that these inhibitors had no effect on acidic stress–produced p18–IL-1β, whereas KCl, Z-YVAD, and CA074-Me completely or partially inhibited ATP-produced p17–IL-1β (Fig. 4B). To further confirm that NLRP3 inflammasomes are not involved in acidic stress–induced IL-1β production, NLRP3- and caspase-1-deleted THP-1 macrophages were generated by using the CRISPER/Cas9 genome-editing system. As expected, deficiency of NLRP3 and caspase-1 completely inhibited ATP-produced p17–IL-1β, but not severe acidic stress–produced p18–IL-1β (Fig. 4C). These data further support the notion that severe acidic stress produces p18–IL-1β independent of NLRP3 inflammasome.

Although caspase-1 is a predominant protease that processes pro–IL-1β into its mature form, it has been reported that pro–IL-1β can be processed by other proteases, such as proteinase 3, neutrophil elastase, cathepsin G, chymase, and chymotrypsin (32). To determine which protease(s) is important for acidic stress–produced p18–IL-1β, we pretreated THP-1 macrophages with serine or cysteine protease inhibitors such as AEBSF (serine protease inhibitor) and E-64 (cysteine protease inhibitor). Although AEBSF inhibited severe acidic stress–produced p18–IL-1β (Fig. 5A, 5B), E-64 did not (Fig. 5C, 5D). Among serine proteases, cathepsin G is expressed in THP-1 cells (33) and cleaves the N terminus of pro–IL-1β at Tyr113 (32); therefore, we tested the CGI, but it did not affect acidic stress–produced p18–IL-1β (Fig. 5E, 5F).

Because previous studies reported that mild acidic stress (pH 6.2) or lactic acid produces p18–IL-1β in microglia or THP-1 cells (16, 17), we next investigated whether severe acidic stress (pH 1.75)–produced p18–IL-1β would be similar to mild acidic stress (pH 6.2)–produced p20–IL-1β. Consistent with previous reports (16, 17), mild acidic stress (pH 6.2)– or lactic acid–induced IL-1β production, which was significantly inhibited by pretreatment with the cathepsin D inhibitor pepstatin A (Fig. 5G), indicating that mild acidic stress or lactic acid each induced IL-1β processing by cathepsin D. In contrast, neither severe acidic stress (pH 1.75)– nor ATP-induced IL-1β production was inhibited by pepstatin A (Fig. 5H). These findings suggest that severe acidic stress–induced production of p18–IL-1β is mediated by serine protease(s), but not by cathepsin G or cathepsin D in macrophages.

To explore the cleavage site(s) of IL-1β by acidic stress, we constructed several FLAG-tagged human IL-1β mutants. Previous studies identified that pro–IL-1β is recognized by proteases at amino acids E111, Y113, V114, D116, and R120 (32). Of these, E111 is cleaved by the metalloprotease meprin β and the molecular mass of IL-1β cleaved at R120 would be less than that cleaved between D116 and A117; thus, three human IL-1β mutants were constructed: IL-1β_Y113D, IL-1β_V114E, and IL-1β_D116I. The expression levels of each FLAG/pro–IL-1β mutant in THP-1 macrophages were confirmed to be equal (Fig. 6A). Caspase-1 has been shown to cleave pro–IL-1β between D116 and A117. Indeed, ATP treatment processed IL-1β_WT, IL-1β_Y113D, and IL-1β_V114E, but not IL-1β_D116I, into p17–IL-1β (Fig. 6B). In contrast, severe acidic stress processed all of these IL-1β mutants into p18–IL-1β (Fig. 6C).

Because our data suggested that cleavage site(s) of pro–IL-1β by severe acidic stress are different from the sites described previously (Y113/V114, V114/H115, and D116/A117), we performed mass spectrometry analysis to identify a cleavage site processed by acidic stress and identified two candidates: T107/W108 and D109/N110 (Fig. 7A). To identify the responsible cleavage site, we constructed FLAG/pro–IL-1β mutants IL-1β_T107A, IL-1β_W108A, IL-1β_D109A, and IL-1β_N110A. After confirming the equal expression levels of each FLAG/pro–IL-1β mutant in THP-1 macrophages (Fig. 7B), we tested the effect of ATP- or severe acidic stress–induced pro–IL-1β processing of each of these mutants. ATP processed all of these mutants (Fig. 7C, 7D). However, acidic stress processed IL-1β_T107A, IL-1β_W108A, and IL-1β_N110A (Fig. 7E, 7F), but failed to process IL-1β_D109A, indicating that D109 is the cleavage site of pro–IL-1β by acidic stress. Furthermore, through use of the “Compute pI/Mw” tool, the theoretical molecular mass of acidic stress–induced mature IL-1β was calculated to be 18.2 kDa.

Finally, to confirm whether acidic stress–produced p18–IL-1β could exert bioactivity, we examined the effects of the supernatants prepared from acidic stress–treated THP-1 macrophages on p38 phosphorylation in alveolar epithelial A549 cells. The supernatants from the acidic stress condition clearly promoted p38 phosphorylation, which is consistent with previous reports that IL-1β stimulates p38 phosphorylation in A549 cells (Fig. 8A) (34). In addition, p38 phosphorylation was inhibited by pretreatment with IL-1Ra (Fig. 8B). Similarly, the supernatants from the acidic condition significantly induced the expression of chemokines, CCL2 and CXCL1, and this increased expression was significantly inhibited by IL-1Ra (Fig. 8C). These results indicate that acidic stress–produced p18–IL-1β exerts bioactivity as its mature form.

The major findings of this study are as follows: 1) acid aspiration induced inflammatory responses, ROS generation, and ALI in WT mice, and these effects were significantly attenuated in IL-1β^−/− mice; 2) acid aspiration–induced ALI and IL-1β production were not inhibited in NLRP3^−/− mice; 3) in vitro experiments revealed that severe acidic stress (pH 1.75; similar to gastric fluid) processed pro–IL-1β into p18–IL-1β, which was different from p17–IL-1β processed by caspase-1, in human THP-1 macrophages and primary murine macrophages; 4) deficiency of NLRP3 and caspase-1 had no effect on acidic stress–produced p18–IL-1β in macrophages; 5) acidic stress–induced production of p18–IL-1β was prevented by inhibitors for serine proteases (AEBSF), but not cysteine proteases (E-64), cathepsin G, cathepsin D, or NLRP3 inflammasome; 6) acidic stress processed pro–IL-1β with known cleavage site mutations (IL-1β_Y113D, IL-1β_V114E, and IL-1β_D116I); 7) using mass spectrometry and processing-site mutants of pro–IL-1β, we identified D109/110 as a novel cleavage site of pro–IL-1β in response to severe acidic stress; and 8) the bioactivity of acidic stress–produced p18–IL-1β was confirmed by its ability to induce p38 phosphorylation and chemokine upregulation in alveolar epithelial cells. These results clearly indicate that acidic stress processes pro–IL-1β at a site different from that previously reported and promotes IL-1β–driven inflammation independent of NLRP3 inflammasome, and contributes to the development of acid aspiration–induced ALI. Our study also shows a novel mechanism of acid-induced IL-1β production and inflammation, and provides new insight into the therapeutic strategies for gastric aspiration pneumonitis and ALI.

Increasing evidence indicates that sterile inflammation contributes to the development of aspiration pneumonitis and ALI (1, 2). Indeed, several inflammatory cytokines and chemokines including TNF-α and MIP-2 have been shown to be involved in its pathogenesis (35, 36). Although IL-1β plays a pivotal role in the process of sterile inflammation, its role in gastric aspiration–induced pneumonitis and ALI has not yet been demonstrated. In the current study, we clearly showed that IL-1β levels were elevated after acid aspiration and that IL-1β deficiency significantly improved acid aspiration–induced inflammatory responses and ALI, indicating that IL-1β is required for its pathogenesis. Regarding IL-1β production, NLRP3 inflammasome has received much attention because it contributes to the development of sterile inflammatory diseases (5, 6). Therefore, we hypothesized that NLRP3 inflammasome could mediate acid aspiration–induced IL-1β production in the lung. Contrary to our expectation, deficiency of NLRP3 failed to prevent acid aspiration–induced inflammatory responses and ALI. Furthermore, acid aspiration–induced IL-1β production was not decreased by NLRP3 deficiency. Previous studies have suggested that NLRP3 inflammasome plays a key role in several ALI models, such as LPS alone and combined with mechanical ventilation (37, 38). Grailer et al. (39) also reported that extracellular histones released from neutrophils activates NLRP3 inflammasome and induces inflammatory responses and lung injury in C5a- and IgG immune complex–induced ALI. In contrast, we have recently reported that hyperoxia induces ALI dependent of NLRP3, but independent of IL-1β (22). Furthermore, Cheng et al. (40) recently highlighted the role of caspase-11 noncanonical inflammasome in a murine model of LPS-induced ALI. Therefore, we postulate that the contribution of NLRP3 inflammasome or IL-1β may depend on the ALI model employed. To our knowledge, this study is the first to demonstrate that NLRP3 inflammasome–independent IL-1β production contributes to the development of aspiration pneumonitis and subsequent ALI.

Homeostasis of cellular pH is essential for the proper cell function and stability. Previous studies have shown that changes in the cellular pH environment influence inflammatory responses (15, 17, 25, 26). With respect to IL-1β, two studies have reported that cathepsin D–dependent production of p20–IL-1β occurs under mild acidic conditions, and suggested that this p20–IL-1β is distinct from p17–IL-1β processed by caspase-1 (16, 17). In these studies, however, cells were subjected to mild acidic conditions (pH 6–7) because tissue ischemia has been shown to drop at approximately pH 6–6.5 (13, 14). Therefore, IL-1β production under severe acidic conditions similar to the gastric environment has not yet been clarified. In the current study, we showed that acidic stress (pH 1.75) induced the production of p18–IL-1β, independent of NLRP3 inflammasome, in macrophages. Interestingly, the well-known NLRP3 inflammasome activator ATP produced not only p17–IL-1β, but also p18–IL-1β; the amount of p18–IL-1β was much less than that of p17–IL-1β. In contrast, although acidic stress produced both p17–IL-1β and p18–IL-1β, the large majority was p18–IL-1β. These results are partly consistent with previous reports that danger signals, such as ATP, monosodium urate, and calcium pyrophosphate dehydrate, produce p20–IL-1β under mild acidic conditions in microglial cells (16, 17). In addition, we showed that severe acidic stress produced p18–IL-1β in a serine protease–dependent, but cathepsin D–independent, manner, whereas mild acidic stress produced p20–IL-1β in a cathepsin D–dependent manner. Therefore, we assumed that although severe and mild acidic stress produce the mature form of IL-1β with similar molecular mass, the responsible protease(s) differs depending on the pH level. Because V113, Y114, and D116 are known to be cleaved by proteinases (neutrophil elastase, cathepsin G, chymase, and chymotrypsin proteinase 3 for V113; proteinase 3 for Y114; and meprin A/β for D116) (32), we tested whether acidic stress–induced processing could be inhibited in IL-1β_Y113D, IL-1β_V114E, and IL-1β_D116I; however, severe acidic stress processed all of these IL-1β mutants into p18–IL-1β, indicating that acidic stress–produced p18–IL-1β is cleaved at sites other than Y113, V114, and D116. The results prompted us to perform mass spectrometry, and subsequent analysis using cleavage mutations of pro–IL-1β revealed that D109 is a novel cleavage site of pro–IL-1β in response to acidic stress. We further calculated that the theoretical molecular mass of acidic stress–induced mature IL-1β was 18.2 kDa. These findings suggest that this cleavage site is a novel potential therapeutic target for aspiration pneumonitis.

Several limitations of this study should be noted. First, we showed that macrophages are the predominant effector cells that produce IL-1β under severe acidic conditions because alveolar macrophages are directly exposed to acidic stress in the setting of gastric aspiration; however, the role of other cell types (e.g., fibroblasts, epithelial cells, and endothelial cells) in the acidic stress–induced production of IL-1β remains to be determined. Second, we showed that severe acidic stress–induced p18–IL-1β clearly induces p38 phosphorylation and chemokine expression, indicating its bioactivity. Edye et al. (26) recently reported that mild acidic stress–induced p20–IL-1β was minimally active in the transduction of IL-1 signaling and suggested that cathepsin D–mediated p20–IL-1β may act as a negative regulator of p17–IL-1β that limits the amount of pro–IL-1β available for caspase-1 processing. Third, although several interventions including platelet depletion and IL-17Ra deletion have been shown to prevent the development of acid aspiration–induced ALI (41, 42), there does not seem to be specific clinical intervention that is accepted for the treatment of acid aspiration–induced ALI. In this regard, gastric fluid commonly contains particulate food matter in the clinical setting, and several experimental studies have reported that aspiration of acidic stress with gastric particles synergistically exacerbated acid aspiration–induced ALI (2, 35). Thus, further investigations are necessary to elucidate the mechanism and role of IL-1β in the pathophysiology of aspiration pneumonitis.

In conclusion, we demonstrated that NLRP3 inflammasome–independent IL-1β production contributes to the development of gastric aspiration pneumonitis. We also showed that severe acidic stress induces the serine protease–dependent production of mature IL-1β in macrophages and identified a novel cleavage site of IL-1β in response to severe acidic stress. The findings of the current study demonstrate that IL-1β plays a pivotal role in acid aspiration–induced inflammation and ALI and identify the novel mechanism underlying acidic stress–mediated IL-1β processing. Furthermore, this study provides new insight into therapeutic strategies for aspiration pneumonitis.

We thank Dr. Tsukasa Ohmori (Jichi Medical University) for invaluable suggestions, and Dr. Dirk Bohmann (University of Rochester, Rochester, NY) for providing the plasmids.

The authors have no financial conflicts of interest.