Occupational exposure to crystalline silica is associated with the development of pulmonary inflammation and silicosis, yet how silica initiates pulmonary fibrosis and which cell types are involved are unclear. In studies here, we hypothesized that silica particles interact initially with pulmonary epithelial cells and alveolar macrophages (AMs) to cause transcriptional activation of nuclear factor (NF)-κB-regulated genes encoding inflammatory cytokines. Exposure of NF-κB luciferase reporter mice intratracheally to silica or lipopolysaccharide (LPS), but not the nonfibrogenic particle titanium dioxide (TiO2), increased immunoreactivity of luciferase protein in bronchiolar epithelial cells and AMs. Ribonuclease protection assays revealed significant (P ≤ 0.05) increases in mRNA levels of inducible nitric oxide synthase, tumor necrosis factor-α, macrophage inflammatory protein-2, macrophage chemotactic protein-1 (MCP-1), interferon-γ, interleukin (IL)-6, and IL-12 in lung homogenates of reporter mice after exposures to silica or LPS. Immunoreactivity of MCP-1 in these animals was localized to AMs and epithelial cells. These data are the first to show activation of NF-κB in situ by fibrogenic particles in pulmonary epithelial cells and AMs. Increased expression of NF-κB-related inflammatory cytokines by these cell types, which first encounter silica after inhalation, may be critical to the initiation of silica-associated lung diseases, thus providing a rationale for focusing on NF-κB in preventive and therapeutic strategies.
- transgenic mice
- nuclear factor-κB
crystalline silica is associated in the workplace with the development of silicosis (1) and recently has been categorized as a human carcinogen in lung (21). Although the mechanisms of pulmonary fibrosis by silica are under investigation in a number of laboratories (reviewed in Ref. 28), the critical events involved in initiation of pulmonary inflammation and lung disease are obscure. In human and rodent models, exposure to silica particles elicits a significant and sustained inflammatory response characterized by the influx of inflammatory cells (8, 16) and increased expression of inflammatory cytokines such as tumor necrosis factor (TNF)-α (19, 29), interleukin (IL)-1 (35), inducible nitric oxide synthase (iNOS) (4, 33), and interferon (IFN)-γ (12, 13). Regulation of these cytokines may be dependent upon the activation of the transcription factor, nuclear factor (NF)-κB, because the promoter regions of many of these genes are known to contain binding sites for this transcription factor (5).
NF-κB is a ubiquitous transcription factor that can be activated by cytokines, reactive oxygen species (ROS), growth factors, bacteria and viruses, ultraviolet irradiation, and inorganic particles (reviewed in Ref. 25). The regulation of NF-κB and its degradation are topics of contemporary interest, as many inducible genes that encode cytokines, chemokines, adhesion molecules, growth factors, enzymes, and transcription factors contain binding sites for NF-κB within their promoter or enhancer regions (34). Moreover, modulation of the NF-κB pathway may be critical to treatment of a number of respiratory and other inflammatory diseases in which NF-κB-related genes are upregulated (2).
We and others have shown in a number of in vitro models that the fibrogenic minerals silica and asbestos cause NF-κB activation (9, 10, 22). Moreover, increases in immunoreactivity of p65, a transcriptionally active subunit of NF-κB, have been demonstrated in bronchiolar epithelium and fibrotic lesions of rats after inhalation of asbestos (23). Bronchoalveolar lavage (BAL) cells from rats exposed to silica by intratracheal instillation also showed NF-κB activation as determined by electromobility shift assay (32). However, whether silica causes transcriptional activation of NF-κB-dependent gene expression in the lung after exposures in vivo is unclear. We addressed this question using intratracheal instillation of α-quartz silica into NF-κB luciferase reporter mice (27) backcrossed into C57BL/6 mice, a strain susceptible to the development of fibrotic lung disease by asbestos or silica. This protocol gives rise to pulmonary inflammation and silicosis (8). In addition, we used ribonuclease protection assays (RPA) to determine whether mRNA levels of genes encoding specific inflammatory cytokines were increased in the lung after silica exposures and immunocytochemistry to determine the cell types involved in NF-κB transactivation and cytokine expression. The specificity of these responses was determined using a known inflammatory agent, lipopolysaccharide (LPS), and a nonpathogenic control particle, TiO2 (15, 24).
NF-κB luciferase reporter mice.
The NF-κB luciferase reporter mice were made using the pBIIX-luciferase construct with two copies of the κB sequence from the Igκ intronic enhancer and were characterized as described previously (27). Thus increases in luciferase protein expression demonstrate functional NF-κB activation of the luciferase reporter construct, a tandem repeat of the canonical NF-κB response element found in the Igκ enhancer. The transcript was expressed in all cell types (27). Mice were backcrossed 4× onto a C57BL/6 background, a strain exhibiting inflammation and fibrosis in response to silica or asbestos (28). Animals were housed at the University of Vermont under controlled conditions of temperature, humidity, and light and provided food and water ad libitum. Animal facilities are American Association for Accreditation of Laboratory Animal Care approved and operated under the supervision of the Institutional Animal Care and Use Committee of the University of Vermont.
Exposures to particles.
Silica particles (α-quartz; Min-U-Sil 5; 0.6 μm, mean equivalent spherical diameter; Pennsylvania Glass and Sand, Pittsburgh, PA) were administered to mice as previously described (8). Briefly, mice were anesthetized with pentobarbital sodium (40 mg/kg ip), and their tracheas were exposed by dissection. Silica particles were instilled intratracheally at 1 mg/0.05 ml sterile Ca2+-Mg2+-free phosphate-buffered saline (CMF-PBS)/mouse. LPS (026:86; Sigma, St. Louis, MO) was instilled at 2 μg/mouse (100 μg/kg) in 0.05 ml sterile CMF-PBS. TiO2(fine, 0.25-μm mean equivalent spherical diameter; obtained from Dr. Gunter Oberdörster, University of Rochester, Rochester, NY), a nonpathogenic control particle, was instilled at 1 mg/mouse in 0.05 ml of sterile CMF-PBS. Vehicle control animals received a single intratracheal injection of 0.05 ml sterile CMF-PBS. Mortality in all groups was <5%.
Animals were killed by an overdose of pentobarbital sodium at 4, 24, or 72 h after exposures to particles, LPS, or PBS (sham vehicle control). Chest cavities were opened, and lungs were cannulated via the trachea with polyethylene tubing. Lungs were then lavaged in situ six times with CMF-PBS at a volume of 1 ml for each lavage. All lavage (BAL) fluid was centrifuged at 250 g for 10 min at 4°C, and the cell pellets were resuspended in 1% bovine serum albumin (BSA) in CMF-PBS. A portion of these cells was then stained with 1% crystal violet and counted under light microscopy for total cell number. Another portion of the cells was diluted in 1% BSA/CMF-PBS and centrifuged onto slides in a Shandon cytocentrifuge (Shandon Southern Products, Cheshire, England). The cells were either stained with LeukoStat Stain Kit (Fisher Scientific, Pittsburgh, PA) and identified by nuclear morphology or fixed in 2% paraformaldehyde (PFA)/CMF-PBS and stored at −20°C. The remaining cells were evaluated for luciferase enzymatic activity.
Detection of NF-κB activation by immunostaining for luciferase protein.
After lung lavage, two left lobes of the lung were inflated at a constant pressure of 25 cmH2O and preserved in 4% PFA/CMF-PBS. All lungs were embedded in paraffin and sectioned. Lung sections (3 μm thickness) and BAL cells were evaluated for the presence of luciferase protein by immunostaining. Lung sections were deparaffinized in xylene for 5 min 3× and rehydrated through graded ethanols. Both lung sections and fixed BAL cells were equilibrated in CMF-PBS. The sections and cells were boiled in 0.1 M citrate buffer (pH 6.0) for antigen retrieval and then incubated in 3% H2O2 to dampen endogenous peroxidase activity. BAL cells were then permeabilized with 1% SDS. Both lung sections and BAL cells were incubated overnight at 4°C in 2% normal goat serum (Jackson ImmunoResearch Labs, Westgrove, PA) to reduce nonspecific protein binding. The following day, the sections and cells were incubated with rabbit anti-luciferase antibody (0.5 μg/ml) (Cortex Biochem, San Leandro, CA) for 1 h at room temperature (RT). After three washes in CMF-PBS, immunoreactivity was detected using the anti-rabbit IgG Vectastain ABC Elite kit (Vector Laboratories, Burlingame, CA) and diaminobenzidine (DAB) as a chromogen according to the manufacturer's protocols. After color development, lung sections and BAL cells were counterstained with hematoxylin. Lung sections were then dehydrated, cleared, and mounted on slides in VectaMount (Vector Laboratories), and BAL cells were mounted on slides in glycerol gelatin (Sigma) before examination by light microscopy. Negative controls consisted of lung sections incubated with secondary antibody alone or rabbit IgG type-matched monoclonal antibody (Zymed Laboratories, San Francisco, CA) in place of primary antibody.
RPA for inflammatory chemokines and cytokines.
Total RNA was isolated from frozen lavaged lung tissues (right lobe) as described previously (31), quantitated by absorbance at 260 nm, and analyzed using an RPA system with a multiprobe template set for iNOS, IL-12, TNF-α, IL-6, macrophage inflammatory protein (MIP)-2, TGF-β, macrophage chemotactic protein (MCP)-1, and IFN-γ according to manufacturer's instructions (Riboquant; PharMingen, San Diego, CA). RNA duplexes were isolated by extraction/precipitation, dissolved in 5 μl of gel loading buffer, and electrophoresed in standard 5% acrylamide/urea sequencing gels. After gels were dried, autoradiograms were developed and quantitated using a Bio-Rad phosphorimager (Bio-Rad, Hercules, CA). Results were normalized to expression of the housekeeping gene L32.
Immunohistochemistry for MCP-1 protein.
To determine whether elevated mRNA levels of genes encoding inflammatory chemokines and cytokines in lung homogenates reflected increases in protein and the cell types expressing these proteins, immunocytochemistry using an antibody to an NF-κB-regulated inflammatory chemokine, MCP-1, was performed on lung sections. Lung sections were deparaffinized in xylene for 5 min 3×, rehydrated through graded ethanols, and equilibrated in CMF-PBS. The sections were boiled in 0.1 M citrate buffer, pH 6.0, for antigen retrieval and then incubated in 3% H2O2 to dampen endogenous peroxidase activity. Lung sections were incubated overnight at 4°C in 2% normal rabbit serum (Jackson ImmunoResearch Labs) to reduce nonspecific protein binding. The following day, the sections and cells were incubated with goat anti-MCP-1 antibody (2 μg/ml) (Santa Cruz Biotechnology, Santa Cruz, CA) for 1 h at RT. After three washes in CMF-PBS, immunoreactivity was detected using the anti-goat IgG Vectastain ABC Elite kit (Vector Laboratories) and DAB as a chromogen according to the manufacturer's protocols. After color development, lung sections and BAL cells were counterstained with hematoxylin. Lung sections were then dehydrated, cleared, and mounted in VectaMount (Vector Laboratories) before examination by light microscopy. Negative controls consisted of lung sections incubated with secondary antibody alone or rabbit IgG type-matched monoclonal antibody (Zymed Laboratories) in place of primary antibody.
Data are expressed as means ± SE. Differences between groups were evaluated by analysis of variance using the Student-Newman-Keuls procedure to correct for multiple comparisons. A P value of less than or equal to 0.05 was considered significant.
Instillation of silica or LPS causes protracted inflammation in mouse lungs.
Instillation of silica particles (1 mg/C57BL/6 luciferase reporter mice) initiated a prominent inflammatory response as evidenced by increased numbers of polymorphonuclear leukocytes (PMNs) (P ≤ 0.05) as early as 4 h postinstillation (Fig.1, A and B). This cellular inflammatory response was still apparent at 72 h when increases (P ≤ 0.05) in total cell numbers also were observed. Exposure of mice to LPS (2 μg) also elicited a marked inflammatory response comprising increased total cell numbers and elevations in the proportions of PMNs and alveolar macrophages (AMs) at all time points. Instillation of the nonfibrogenic particle TiO2 did not elicit significant increases in total cell numbers or percentages of PMNs.
Luciferase protein is increased in AMs and bronchiolar epithelial cells in luciferase reporter mice exposed to silica or LPS.
NF-κB reporter mice were also assessed for luciferase activity as an indication of transcriptional activation of NF-κB gene expression (27). Homogenized whole lung tissue after exposures to agents did not demonstrate significantly increased enzymatic activity above that seen in PBS-instilled mice (data not shown), presumably because epithelial cells and AMs are only a small fraction of lung tissue. However, examination of lung tissue and BAL cells by immunohistochemistry revealed focally increased luciferase immunoreactivity in both AMs and epithelial cells of silica- and LPS-instilled mice (Fig. 2). Compared with PBS controls (Fig. 2 A), exposure of mice to LPS (Fig.2 B), an agent activating NF-κB in lung (6), or to silica for 24 or 72 h (Fig. 2, C andD) elicited increases in luciferase protein in both AMs (arrow) and in bronchiolar epithelial cells. At 72 h postinstillation, less intense staining was detected in lung sections from mice exposed to LPS (data not shown). Increases in staining were not observed in alveolar epithelial cells from asbestos or LPS-exposed mice. No staining was seen in lung tissue from mice exposed to TiO2 despite its accumulation in lung (Fig. 2 E, arrow) or in silica-exposed lungs stained with secondary antibody alone (Fig. 2 F). Moreover, no staining was observed in lung tissues from wild-type C57BL/6 mice.
Instillation of silica or LPS causes an increase in mRNA levels of NF-κB-related inflammatory cytokines.
Possible consequences of NF-κB activation after exposure to silica or LPS were investigated by examining lung homogenates for increased mRNA levels of a number of NF-κB-associated or -regulated inflammatory cytokines in C57BL/6 wild-type mice. Figure3 is a representative autoradiograph from an RPA on lung homogenates of mice injected with PBS, LPS, or silica. Figure 4 shows quantitation by phosphorimaging of steady-state mRNA levels of these cytokines in lung homogenates of mice at 4 and 24 h after instillation of agents. Compared with sham (PBS) mice, instillation of silica elicited a significant (P ≤ 0.05) increase in mRNA of several inflammatory cytokines (iNOS, MIP-2, MCP-1, IFN-γ, IL-12, and IL-6) as early as 4 h postinstillation (Fig. 4 A). Significant increases in mRNA levels of TNF-α, iNOS, MIP-2, MCP-1, and IFN-γ also occurred at 24 h (Fig. 4 B), whereas only mRNA levels of IFN-γ were increased at 72 h postinstillation of silica (data not shown). LPS caused elevations in mRNA levels of all cytokines at 4 h, which persisted, with the exception of TGF-β1, for 24 h. Increases in mRNA levels of cytokines were not observed with instillation of TiO2 (data not shown) and were identical to saline-instilled controls.
MCP-1 protein is increased in AMs and epithelial cells in mice exposed to silica.
To document whether increased mRNA levels of an NF-κB-regulated gene, MCP-1 (38), reflected increased protein and the cell types involved, lung sections from all groups were evaluated for MCP-1 protein using immunohistochemistry. As shown in Fig.5, sham mice instilled with PBS demonstrated some constitutive expression of MCP-1 in bronchiolar epithelial cells (Fig. 5 A). Whereas LPS increased MCP-1 immunoreactivity primarily in AMs (Fig. 5 D), silica elicited marked increases in immunoreactivity of MCP-1 in AMs and bronchiolar epithelium (Fig. 5, E and F). No staining was found in lung tissue from mice exposed to secondary antibody alone (Fig. 5 B).
Silicosis is associated with occupational exposures to crystalline silica in stone cutting, quarrying, and mining (1, 21). Increases in pulmonary inflammation may be related to the development of fibrotic lung disease, although the critical molecular events involved in initiation of inflammation are unclear. Work here demonstrates for the first time that fibrogenic silica particles in vivo activate the transcription factor NF-κB in AMs and bronchiolar epithelial cells of the lung. The epithelial cell specificity of transactivation is supported by our previous work showing increased localization of p65 protein in bronchiolar epithelium of mice exposed to asbestos by inhalation (23). NF-κB transactivation may participate in the regulation of key inflammatory mediators and the development of inflammation and fibrosis, features of these model systems. Although trends were observed, significant increases in luciferase enzymatic activity were seen in neither whole lung homogenates nor BAL cell pellets of silica-exposed mice. However, because increased immunoreactivity of luciferase protein occurred only in a small population of cells in the lung, activity may have been diluted beyond the limits of detection in homogenates from an entire lung lobe or in BAL pellets containing many neutrophils.
Activation of NF-κB in epithelial cells or AMs may result in the transcriptional initiation of a diverse set of genes important in perpetuating or attenuating immune and inflammatory responses (reviewed in Ref. 25). Thus steady-state mRNA levels of several inflammatory cytokines and chemokines with NF-κB sites in their promoter regions were measured in lung homogenates of silica-, LPS-, or TiO2-exposed mice. Significant increases in mRNA levels of TNF-α, iNOS, MIP-2, MCP-1, IFN-γ, IL-12, and IL-6 were detected after instillation of silica or LPS, a known inflammatory agent and positive control in our studies.
The elaboration of these cytokines by minerals inducing inflammation and fibrosis may be complex and interrelated (reviewed in Ref.28). For example, intratracheal administration of silica primes AM to release TNF-α after in vitro exposures to LPS (15). Intratracheal instillation of silica also increases TNF-α mRNA levels in lung tissue (30) and release of TNF-α protein from BAL cells (19, 29). After inhalation of silica, persistent overexpression of TNF-α mRNA (measured by in situ hybridization) occurs in alveolar epithelial cells and aggregate lesions in the lungs as well as in mononuclear cells in BAL (12). The critical relevance of TNF-α to fibrosis is demonstrated by experiments in which administration of human recombinant soluble TNF receptor prevents silicosis and bleomycin-induced fibrosis in mice (30).
IFN-γ synergistically enhances TNF-α-induced NF-κB transactivation by a mechanism involving IκB degradation (11). In studies here, elevated and protracted mRNA levels of IFN-γ were observed in lung for as long as 72 h after exposure to silica. Inhalation of silica also elicits overproduction of IFN-γ by lymphocytes in mice developing silicosis (12,13). Moreover, in rats inhaling silica, increased IFN-γ and IL-12 mRNA levels occur primarily in thoracic lymph nodes (18).
Silica also increased mRNA levels of the potent chemokine MIP-2, a factor upregulated by TNF-α and synthesized by bronchiolar and alveolar type II epithelial cells in vitro (14). Increased levels of MIP-2 mRNA are also detected in the lungs and BAL cells of rats after intratracheal instillation of silica (14, 39).
The fact that the epithelial cell may be a major source of cytokines and chemokines after exposures to silica in vivo is supported by our data and several recent reports. In human bronchial epithelial cells (BEAS-2B) cocultured with inorganic particles, increased IL-6 mRNA, a cytokine elevated in BAL fluids from pneumoconiosis patients (37), occurs (17). In addition, increased MCP-1 mRNA levels are seen in a mouse type II epithelial cell line (MLE) after exposure in vitro to silica (3).
We document here increased immunoreactivity of MCP-1 in both AMs and epithelial cells of silica-exposed mice at sites corresponding to transactivation of NF-κB-dependent gene expression (Figs. 3, 6). MCP-1, an NF-κB-regulated gene, is produced by fibroblasts, macrophages, and epithelial cells in vitro and is a potent chemotactic factor for circulating monocytes (26). The possibility that this chemokine participates in the pathogenesis of particle-induced lung injury is supported by studies showing increased levels of soluble MCP-1 in BAL cells and its increased immunoreactivity in alveolar type II epithelial cells in lung sections from patients with coal worker's pneumoconiosis (7). MCP-1 has also been implicated in fibrosis via its regulation of profibrotic cytokine generation and matrix deposition. For example, T helper (Th) 2 type fibroblasts from murine fibrotic pulmonary granulomas generate twofold more MCP-1 than Th1 type fibroblasts from nonfibrotic lungs (20).
In summary, our data are the first to show transactivation of NF-κB-dependent gene expression in vivo by fibrogenic particles in bronchiolar epithelial cells and AMs. These changes were observed after instillation of LPS, but not in response to TiO2, a nonfibrogenic particle (24), and were accompanied by increased mRNA levels of NF-κB-related cytokines and chemokines in lung tissue. A key NF-κB-regulated cytokine, MCP-1, was localized in epithelial cells and AMs corresponding to sites of NF-κB transactivation. Further studies using transgenic mice in which NF-κB is inhibited in bronchiolar epithelial cells (30a) will be necessary to determine a cause-effect relationship. Increased expression of NF-κB-related inflammatory cytokines by epithelial cells and macrophages may initiate critical molecular events leading to the development of fibrotic lung disease, suggesting modulation of this transcription factor in approaches to prevent and treat pulmonary fibrosis.
We acknowledge the valuable technical assistance of Ingrid Berlanger and Andrew Cummins and the secretarial assistance of Laurie Sabens. Dr. Douglas Taatjes provided expertise on cell imaging approaches and photomicroscopy.
This work was supported by National Institutes of Health Grants RO1 HL-39469 and RO1 ES/HL-09213.
Address for reprint requests and other correspondence: B. T. Mossman, Dept. of Pathology, Univ. of Vermont College of Medicine, Burlington, VT 05405 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2002 the American Physiological Society