Regulatory role of heme oxygenase-1 in silica-induced lung injury

Male BALB/c mice (6 weeks of age) were purchased from Japan SLC (Shizuoka, Japan), housed in light- and temperature-controlled rooms, and given free access to tap water and commercial laboratory chow. The mice were used for experiments following a one-week acclimation period.

The mice were anesthetized with ketamine (80 mg/kg; Sigma-Aldrich, St. Louis, MO, USA) or sodium thiopental (50 mg/kg Nembutal; Dainippon-Sumitomo Seiyaku. Co., Ltd., Osaka, Japan) and xylazine (10 mg/kg; Sigma). A 22-gauge cannula (Terumo, Tokyo, Japan) was inserted through the orotracheal route. Sterilized crystalline silica (Min-U-Sil-5, 100 mg/kg; US Silica, Berkeley Springs, WV, USA) in 100 μl of sterile saline was instilled into the trachea [21]. To induce HO-1 gene expression, 100 μmol/kg hemin (Sigma) was administered intraperitoneally 48, 24, and 0.5 h before silica administration. To inhibit ERK or HO-1 enzyme activity, U0126 (30 mg/kg; Merck, Darmstadt, Germany) or HO-1 inhibitors of either zinc protoporphyrin (ZnPP, 100 μmol/kg; Porphyrin Products, Logan, UT, USA) as a competitive HO-1 inhibitor or ketoconazole (KTZ, 100 μmol/kg; Sigma) as a selective HO-1 inhibitor was administered intraperitoneally in the same manner, respectively [22‐25]. The lungs were removed 1, 2, 3, 7, and 14 days after silica instillation.

The mouse macrophage-like cell line RAW264.7 (derived from BALB/c mice) from the American Type Culture Collection (ATCC; Rockville, MD, USA), and bronchial epithelial cells, transformed human bronchial epithelial cells 16HBE, kindly provided by D.C. Gruenert (Gene Therapy Center, University of California, CA, USA) were used [26]. The cells were cultured in DMEM (Sigma) supplemented with 10% FCS (Equitech-Bio, Kerrville, TX, USA), 1% penicillin, and 1% streptomycin (Sigma) at 37 °C in 5% CO₂. The cells were cultured to subconfluence in 6-well plates (Sumitomo, Osaka, Japan), then rendered quiescent in medium containing 0.5% FCS for another 1 day, followed by exposure of 0.1 or 0.5 mg/ml of sterilized silica for 24 h. In some experiments, the cells were pretreated with one of the following agents before silica treatment: 1 h with 50 or 100 μM of the ERK inhibitor, U0126 (Cell Signaling Technology, Danvers, MA, USA), 1 h with 25 mM or 50 mM tetramethylthiourea (TMTU; MP Biomedicals, Santa Ana, CA, USA), 16 h with 5 or 10 μM bilirubin (Sigma), 1 h with 20 or 50 μM RuCO (CO-releasing molecule; Sigma), 1 h with 100 or 200 μM hemin, or 1 h with 100 or 200 μM ZnPP. The cells were collected either 2–6 or 8–24 h after silica exposure and evaluated for ERK activation or HO-1 induction, respectively, in RAW264.7 or 16HBE cells.

The lung was homogenized in lysis buffer containing 0.25 M sucrose, 20 mM Tris-HCl (pH 7.4), and protease inhibitor (Sigma). After centrifugation (15,000×g for 10 min), the supernatants were collected as the cytoplasm fraction. The pellets were sonicated and further separated into the nuclear, mitochondrial, and microsomal fractions by multi-step centrifugations as previously described [27]. The nuclear fractions were resolved with nuclear extraction buffer containing 20 mM HEPES (pH 7.6), 20% glycerol, 500 mM NaCl, 1.5 mM MgCl₂, 0.2 mM EDTA (pH 8.0), 1 mM DTT, 0.1% NP-40, and protease inhibitor. The protein concentration in each sample was determined by the Bradford method using the Bio-Rad Protein assay kit (Bio-Rad, Hercules, CA, USA). After 5 min boiling of 10 μg microsomal extraction or 20 μg nuclear extraction with an equal volume of sample buffer, the samples were run on 4–20% gradient polyacrylamide gels (Daiichi Kagaku, Tokyo, Japan) for electrophoresis. Then, the proteins in the membranes were transferred onto Immobilon polyvinylidene difluoride membranes (Millipore, St. Louis, MO, USA). The membranes were blocked in blocking buffer (5% nonfat dry milk in Tris-buffered saline, pH 7.5, 0.05% (v/v) Tween20 (TBST)) for 1 h at room temperature, and then incubated with the relevant antibodies for 1 h at room temperature. The following rabbit polyclonal antibodies were used: phospho-ERK1/2 (1:1000 dilution), phospho-JNK (1:1000 dilution), phospho-p38 (1:1000 dilution), ERK1/2 (1:1000 dilution) from Cell Signaling Technology, and HO-1 (OSA-111; ENZO Life Sciences, Farmingdale, NY, USA; 1:1000 dilution). After washing with TBST, the blots were incubated with the appropriate peroxidase-conjugated secondary antibody (1:5000 dilution) for 1 h at room temperature and developed using an enhanced chemiluminescence system (ECL; Amersham, Buck, UK) and Hyperfilm (Amersham). The blotted protein was quantified densitometrically with ImageJ image analysis and image processing software (Version 1.51; National Institutes of Health Image Engineering, Bethesda, MD, USA). β-Actin (A1978; Sigma; 1:10000 dilution) was used as a positive control.

To confirm the effect of silica on HO-1 activation, we studied a murine model of silicosis. After acclimation, crystalline silica was administered intratracheally to BALB/c mice under anesthesia. Immunoblotting analysis was performed on lung homogenates of silicosis mice. The expression levels of HO-1 protein in the lungs were significantly increased 2 days after silica instillation compared with those from control mice (Fig. 1a/b). These observations are consistent with the previous findings [21].

MAPK systems are known to be major factors affecting the disease progression of silicosis [7, 28], with HO-1 as a newly recognized factor [20, 21]. However, the relationship between MAPK systems and HO-1 after silica exposure has not yet been elucidated. To investigate the key signaling pathway involved in the HO-1-mediated response to silica exposure, we first examined phosphorylated MAPK proteins of ERK, p38, and JNK in lungs from murine silicosis. Mice were divided into four groups: 1) pretreated with hemin, an inducer of HO-1, then treated with silica, 2) pretreated with ZnPP, a competitive inhibitor of HO-1, then treated with silica, 3) treated with silica alone, and 4) treated with saline alone. Silica-induced MAPK activation was examined and compared with or without pretreatment with the HO-1 inducer and inhibitor. As shown in Fig. 2a, expression levels of phosphorylated ERK in the lungs were upregulated 1 day after silica exposure, and then gradually decreased. In contrast, expression levels of phosphorylated p38 and JNK were continually increased after silica exposure and were not altered with or without pretreatments of either hemin or ZnPP (Fig. 2a). Most importantly, the expression level of phosphorylated ERK was significantly decreased by pretreatment with hemin, but was significantly increased by pretreatment with ZnPP (Fig. 2a/b). These results suggest that the beneficial effects of HO-1 induction in murine silicosis are associated with the ERK signaling pathway. Consistent with this, mice subjected to intraperitoneal administration of the ERK inhibitor, U0126, 2 h before and 6 h after silica administration showed attenuated HO-1 induction in the lungs (Additional file 1). These results indicated the feedback system between HO-1 and ERK activation. To determine the specific involvement of HO-1, further experiments using KTZ as a selective inhibitor of HO-1 were performed. Mice were administered KTZ intraperitoneally 48, 24, and 0.5 h before silica administration. As shown in Fig. 3a/b, pretreatment with KTZ significantly inhibited HO-1 induction in the lungs 2 days after silica instillation, and the level of activated ERK was significantly higher in silicosis mice pretreated with KTZ compared to those without (Fig. 3c/d). These results are comparable to those using the competitive HO-1 inhibitor ZnPP (Fig. 2). Taken together, these data suggested that HO-1, rather than another heme oxygenase, negatively regulates phosphorylation of ERK.

As shown in our previous report, silica-induced HO-1 upregulation was evident in macrophages and bronchial epithelial cells in silicotic lungs from both mice and humans [21]. Thus, to further investigate the relationship between HO-1 and its products with the ERK pathway, the protein expression of phosphorylated ERK and HO-1 in silica-stimulated RAW264.7, a macrophage cell line (Fig. 4a), and 16HBE, a bronchial epithelial cell line (Fig. 4b), was analyzed by immunoblotting. As shown in Fig. 4a/b, phosphorylated ERK expression, indicating ERK activation, was obvious when more than 30 μg or 300 μg of silica was administered in RAW264.7 or 16HBE cells, respectively. HO-1 expressions were increased in a dose-dependent manner in both cell lines exposed to silica (Fig. 4a/b). For subsequent studies, the dose of silica was set at 100 or 500 μg for either RAW264.7 or 16HBE cells, concentrations which produced significant ERK activation and HO-1 induction, respectively (Fig. 4a/b lower panel).

Since ROS are thought to play a major role in the pathogenicity of crystalline silica [8], we next evaluated the relationship between ROS and HO-1. Cells were stimulated with silica in the presence of a hydroxyl radical scavenger, TMTU, and ERK inhibitor, U0126. As shown in Fig. 5, either TMTU or U0126 dose-dependently suppressed not only phosphorylation of ERK but also HO-1 induction in both RAW264.7 (Fig. 5a) and 16HBE (Fig. 5b) cells exposed to silica. These data suggest that ROS blocking by the radical scavenger led to the suppression of ERK activation, resulting in decreased HO-1 expression.

As shown in the result from the silicosis model (Fig. 2), pretreatment with hemin could effectively pre-induce HO-1 in the lungs, whereas pretreatment with ZnPP suppressed HO-1 induction after silica administration. Similarly, Fig. 6a/b shows expression levels of phosphorylated ERK were suppressed by pretreatment with hemin, but not with ZnPP in both RAW264.7 and 16HBE cells. These results indicated that HO-1 negatively regulates phosphorylation of ERK, which is in agreement with that observed in murine silicosis (Fig. 2).

Based on the results so far, the ROS-ERK pathway could be a key mediator of silica-induced HO-1 upregulation. Finally, we further investigated which of the heme degradation products (bilirubin or CO) was responsible for regulation of the ERK pathway, because hemin-mediated pre-induction of HO-1 could suppress ERK activation (Figs. 2 and 6a/b). Cells were exposed to bilirubin, an antioxidant, or RuCO, a CO-releasing molecule, before silica treatment, and the results showed that CO especially suppressed phosphorylated ERK expression (Fig. 7a/b). These results are consistent with previous reports describing the effect of HO-1 (and its by-product) on modulating ERK [29‐31]. Interestingly, HO-1 expression was associated dose-dependently with pretreatment of either bilirubin or RuCO (Fig. 7a/b). As the ERK pathway has been shown to be involved in the transcription of HO-1 [32], the present results indicate that the exposure to silica particles triggered hydroxyl radical generation and induced HO-1 through activation of the ERK pathway (Figs. 4a/b and 5a/b). Furthermore, high HO-1 expression could regulate silica-mediated ERK activation by heme degradation products, CO and bilirubin (Figs. 6a/b and 7a/b).