Crystalline and amorphous silica differentially regulate the cyclooxygenase-prostaglandin pathway in pulmonary fibroblasts: implications for pulmonary fibrosis

Katherine M. A. O'Reilly, Richard P. Phipps, Thomas H. Thatcher, Beth A. Graf, John Van Kirk, Patricia J. Sime


Inhalation of crystalline (CS) and amorphous silica (AS) results in human pulmonary inflammation. However, silicosis develops only following CS exposure, and the pathogenic mechanisms are poorly understood. This report describes the differential abilities of CS and AS to directly upregulate the early inflammatory mediator COX-2, the recently identified prostaglandin E (PGE) synthase and the downstream mediator PGE2 in primary human lung fibroblasts. Increased cyclooxygenase (COX)-2 gene transcription and protein production were demonstrated by ribonuclease protection assay, Western blot analysis, and immunocytochemistry. In each case the ability of AS to induce COX-2 exceeded that of CS. Similarly, downstream of COX-2, production of the antifibrotic prostaglandin PGE2 was induced in a dose-dependent fashion, but AS was significantly more potent (maximal production: CS = 4,710 pg/ml and AS = 7,651 pg/ml). These increases in COX-2 and PGE2 were preceded by induction of the PGE2 synthase protein, demonstrating the potential role of this novel molecule in silica-mediated inflammation. There was specificity of induction of prostaglandins, as PGF, but not PGD2, was induced. Using specific COX-2 inhibitors, we showed increased PG production to be dependent on the COX-2 enzyme. Furthermore, stimulation of fibroblasts was particle specific, as silica but not carbon black resulted in fibroblast activation. These results demonstrate that silica can directly stimulate human lung fibroblasts to produce key inflammatory enzymes and prostaglandins. Moreover, they suggest a mechanism to explain the differing fibrogenic potential of CS and AS. The molecules COX-2, PGE synthase, and PGE2 are identified as effectors in silicosis.

  • silicosis
  • prostaglandin E2
  • prostaglandin E synthase
  • inflammation

silicosis is an important occupational lung disease that results from the chronic inhalation of crystalline silica (5, 22). Pathologically, it is characterized by the development of inflammation, silicotic nodules, and interstitial scarring. This scarring or fibrosis leads to distortion of the normal lung architecture and disruption of the normal physiological processes of the lung.

There are two distinct types of silica found in nature, crystalline and amorphous silica. They differ significantly in the injury they cause when inhaled. Both crystalline silica and amorphous silica incite pulmonary inflammation, but significantly, only inhalation of crystalline silica ultimately leads to the development of pulmonary fibrosis. It is unclear why there is such a difference in the pathological outcome. Some investigators have suggested that the difference may be related to the physical properties of amorphous and crystalline silica (13, 30). When inhaled crystalline silica particles are ingested by alveolar macrophages, the macrophages are activated and damaged, resulting in ongoing injury and ultimately leading to the development of fibrosis (4). In contrast, the amorphous form has a far larger surface area and hence greater relative solubility within the lung, enabling it to be cleared more easily following inhalation. However, to date little attention has focused on the individual abilities of the crystalline and amorphous forms to induce potentially antifibrotic molecules.

Much of the work on the pathogenesis of silicosis has focused on the role of the alveolar macrophage and, to a lesser extent, the alveolar epithelial cell. Certainly, both alveolar macrophages and epithelial cells secrete proinflammatory mediators following exposure to silica (21, 28, 32, 39). However, despite the close proximity of fibroblasts to inhaled crystalline silica particles when the epithelial layer has been denuded, the contribution of fibroblasts to silica-induced inflammation and fibrosis has not been well studied (32a).

The crucial role of the fibroblast in modulating immune-inflammatory reactions in a number of different tissues and disease states is being increasing recognized (14, 34, 40). When activated, fibroblasts are capable of producing a diverse array of inflammatory mediators, including interleukin-8 (IL-8), monocyte chemoattractant protein-1 (MCP-1), IL-6, cyclooxygenase-2 (COX-2), and transforming growth factor (TGF)-β. Despite this recognition, however, the possibility that fibroblasts can be directly activated by exposure to particulates like silica has received little, if any, attention. This is an area of key importance as fibroblasts may be capable of directing the tissues’ response to injury, either promoting resolution or propelling a cycle of ongoing scarring.

Prostaglandin E2 (PGE2) is one of the key prostaglandins produced at sites of inflammation and fibrosis. It is secreted by resident fibroblasts, as well as inflammatory cells, in response to early warning signals like TNF-α and IL-1β (8). The prostaglandin synthesis pathway begins with the release of arachidonic acid from phospholipid membranes by the action of phospholipase A2 (PLA2). Arachidonic acid is the substrate for two distinct enzymatic pathways, cyclooxygenase (COX) and 5-lipoxygenase. The end-products of the 5-lipoxygenase pathway are leukotrienes, whereas the COX pathway gives rise to prostaglandins and thromboxanes. The COX enzymes convert arachidonic acid to the prostaglandin PGH2, which is further processed to individual prostaglandins by specific prostaglandin synthases. The COX enzyme exists as two isoforms: COX-1, which is constitutively expressed by fibroblasts in all tissues, and the inducible isoform, COX-2 (35). Two forms of PGE synthase have also been described: cytosolic (c) and microsomal (m) PGES. Like COX-1, the cytosolic subtype (cPGES) is generally constitutively expressed, and like COX-2, mPGES is inducible by inflammatory stimuli. mPGES has been shown to be specifically responsible for the upregulation of PGE2 in some models of inflammation (36). To date there has been little study of the role of these enzymes in lung disease and none in dust-induced diseases such as silicosis.

The role of PGE2 in regulating inflammatory responses is complex. It has both proinflammatory as well as antifibrotic properties. PGE2 decreases fibroblast proliferation and collagen gene transcription in vitro (9, 10), and patients with idiopathic pulmonary fibrosis (IPF) have 50% less PGE2 in their bronchoalveolar lavage (BAL) fluid than controls. Similarly, fibroblasts recovered from patients with IPF secrete less PGE2 in response to certain stimuli (16, 37, 38). This strongly suggests that the pathogenesis of this prototypic pulmonary fibrotic disorder, which shares many similarities with silicosis, may be related to an impaired ability to generate PGE2, resulting in the creation of a profibrotic milieu.

This study tests the hypothesis that both crystalline and amorphous silica particles directly activate human lung fibroblasts but differ in their ability to do so. We demonstrate for the first time that both crystalline and amorphous silica can induce production of key enzymes in the prostaglandin pathway and the downstream prostanoids PGE2 and PGF in human primary lung fibroblasts. Interestingly, amorphous silica was found to be considerably more potent than crystalline silica in this regard. These data thus further our understanding of the factors at play during the development of silicosis.



Crystalline silica (mean size 0.8 μm), amorphous silica, and carbon black were a generous gift from Dr. David R. Hemenway (University of Vermont). The selective COX-2 inhibitor SC-58125 was purchased from Cayman Chemical (Ann Arbor, MI). Indomethacin was obtained from Sigma Chemical (St. Louis, MO). Monoclonal antibodies against COX-1 and COX-2 for Western blot analysis and immunohistochemistry were purchased from Cayman Chemical. The mouse IgG1 was from Caltag Laboratories (Burlingame, CA). A polyclonal antibody against mPGES was purchased from Cayman Chemical. The reagents for enzyme immunoassay for detection of PGE2 and PGF and the enzyme immunoassay kit for detection of PGD2 were purchased from Cayman Chemical. The IL-1β ELISA was obtained from R&D Systems (Minneapolis, MN).

Primary human lung fibroblast culture conditions.

Normal human primary fibroblast cell strains derived by explant technique were maintained in MEM (Life Technologies, Gaithersburg, MD) supplemented with 10% fetal bovine serum (Sigma Aldrich) and penicillin (100 units/ml), streptomycin (100 μg/ml), and amphotericin (0.25 μg/ml, Life Technologies). These cells were derived from anatomically normal lung from a patient undergoing surgical resection for a benign pulmonary hamartoma. These cells are morphologically consistent with fibroblasts and express the fibroblast markers collagen and vimentin. They do not express CD45, factor VIII, or cytokeratin. Cells were incubated at 37°C in a humidified environment containing 5% CO2 and 95% air. For all experiments, cells were used between passage numbers 2 and 8. Human tissues were obtained under the approval of the University of Rochester Research Subjects Review Board.

Treatment of primary lung fibroblasts with silica.

Before exposure to silica or carbon black, the medium was changed to serum-free MEM for 48 h to reduce serum-induced expression of COX-2 (25).

Before use, silica and carbon black were baked at 180°C for 2 h to remove any trace of lipopolysaccharide (LPS). They were then tested by a limulus assay (sensitivity of 0.6 EU/ml) to confirm that the depyrogenation had successfully eliminated any LPS contamination. Fibroblasts were grown in 96-well plates (104 cells/well) for prostaglandin and cytokine assays, six-well plates (105 cells/well) for Western blot analysis, and 10-cm dishes (5 × 105 cells) for RNA harvest. Suspensions of crystalline and amorphous silica (1–100 μg/ml) and carbon black (100 μg/ml) (an alternate particulate) were made in serum-free MEM. Fibroblast viability at these concentrations was assayed using an 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay. IL-1β (10 ng/ml) that is known to induce COX-2 was used as a positive control. Serum-free MEM was used as the negative control.

RNA isolation and RNase protection assay.

RNA was isolated from the primary fibroblasts 6 h following treatment with TRI reagent (Molecular Research Center, Cincinnati, OH) according to the manufacturer's instructions. Equal amounts of mRNA were used in the RNase protection assays (RPA).

COX-1 and COX-2 antisense probe templates (Cayman Chemical) and a human β-actin template (Ambion, Austin, TX) were used to prepare labeled RNA probes with the MAXIscript in vitro transcription kit (Ambion). Full-length RNA transcripts were gel purified from an 8 M urea acrylamide gel. The RPA was performed using the RPA II kit (Ambion). In brief, RNA was hybridized with labeled probe (8 × 104 cpm) overnight at 42°C. Hybridized samples were digested at 37°C for 30 min with a mixture of RNase A and RNase T1 supplied with the kit. The samples were then separated on a 5% polyacrylamide-8 M urea sequencing gel.


Fibroblasts were seeded (7 × 103 cells/well) onto eight-well chamber slides and grown until subconfluent. They were serum starved as outlined above and, following the addition of silica or carbon black, were incubated under standard conditions for 24 h. The cells were fixed with 4% paraformaldehyde and then immunostained with 10 μg/ml of mouse anti-COX-2 monoclonal antibody or isotype control (mouse IgG1) at 4°C overnight. The following day the secondary antibody (a biotinylated horse anti-mouse antibody) was added followed by a streptavidin-conjugated horseradish peroxidase (HRP; Vector, Burlingame, CA). Immunostaining was visualized with aminoethyl carbazole (Zymed Laboratories, San Francisco, CA).

Western blotting for COX-2 and mPGES.

We harvested protein from confluent fibroblasts monolayers by adding ice-cold RIPA lysis buffer (150 mM NaCl, 50 mM Tris, pH 8.0, 1% NP-40, 1 mM EDTA) with added proteinase inhibitor cocktail (Sigma Aldrich). The protein concentration was quantified by bicinchoninic acid assay (Pierce, Rockford, IL). We solubilized the proteins by them boiling in sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis (PAGE) sample buffer and separated them by SDS-PAGE. Proteins were transferred to nitrocellulose membranes and immunoblotted with the COX-1, COX-2, or mPGES antibodies described above. Proteins were detected with HRP-labeled secondary antibodies and visualized with enhanced chemiluminescence (Perkin Elmer, Boston, MA).

Enzyme immunoassay for prostaglandin production and ELISA for prostaglandins and cytokines.

Lung fibroblasts were seeded onto flat-bottom 96-well plates at a density of 104 cells per well in 200 μl of MEM supplemented as described in the culture conditions above. The cells were incubated until ∼80% confluent. The cells were then serum reduced as previously described. Suspensions of both crystalline and amorphous silica were prepared in serum-free MEM with or without SC-58125 (5 μM) (a specific COX-2 inhibitor) and indomethacin (20 μM) (a nonselective inhibitor that blocks the activity of both COX-1 and COX-2) and incubated for 48 h. The supernatants were assayed for the presence of PGE2, PGF, and PGD2 using specific enzyme immunoassay reagents and kits from Cayman Chemical according to the manufacturer's instructions. IL-1β, MCP-1, IL-6, IL-8, and TGF-β production in the cell supernatants was studied using commercial ELISA kits (R&D Systems) according to the manufacturer's recommendations.


All data are expressed as means ± SE. Statistical significance was assessed using a one-way analysis of variance (ANOVA) test and was recorded at P < 0.05. Where significant differences were found, the Tukey-Kramer multiple-comparisons post hoc test was used to identify differences between individual groups. In cases where SEs were significantly different between groups, a nonparametric ANOVA (Kruskal-Wallis test) was employed, followed by a Dunn's multiple-comparisons post hoc test, to assess statistical significance.


Crystalline and amorphous silica induce COX-2 but not COX-1 in primary human lung fibroblasts.

Primary human lung fibroblasts were treated with increasing concentrations of silica for 24 h, and COX-2 expression was evaluated by Western blot. Both amorphous and crystalline silica induced COX-2 in a dose-dependent manner (Fig. 1). Interestingly, whereas amorphous silica strongly induced COX-2 at 10 μg/ml, 10-fold more crystalline silica was required to observe a similar induction of COX-2 protein. Trypan blue exclusion and MTT assays determined that amorphous or crystalline silica was nontoxic to the fibroblasts at concentrations up to 100 μg/ml (data not shown). Carbon black did not induce COX-2 expression, demonstrating that upregulation of COX-2 is not a nonspecific response to particulates. COX-1 expression was not affected by silica, consistent with previous reports that COX-1 is constitutively expressed and is not induced by inflammatory stimuli. This demonstrates that the induction of COX-2 is a specific proinflammatory response to silica.

Fig. 1.

Silica induces cyclooxygenase (COX)-2 but not COX-1 protein in human lung fibroblasts. Fibroblasts in culture were treated for 24 h with the indicated concentrations of amorphous (AS) or crystalline (CS) silica. Carbon black (CB, 100 μg/ml) was included as a negative control. Total cell protein was analyzed by Western blot for expression of COX-2, COX-1, and actin (to confirm equal protein loading). Amorphous silica induces COX-2 slightly at 2.5 μg/ml and strongly at 10 μg/ml. Tenfold higher concentrations of crystalline silica are required to produce similar elevations in COX-2 protein. Neither form of silica alters the constitutive expression of COX-1. MC, media control.

Immunocytochemical staining was performed to further delineate and localize the expression of COX-2 protein in the fibroblasts. COX-2 expression is markedly upregulated following exposure to crystalline silica (100 μg/ml) and amorphous silica (10 μg/ml) (Fig. 2). Carbon black did not result in any increased expression of COX-2. This suggested that the COX-2 induction observed with both forms of silica is particle specific and is not merely induced by all types of particulates. No staining was observed in untreated cells or any cells stained with the isotype control antibody (data not shown).

Fig. 2.

COX-2 protein expression is induced by silica, but not CB, in human lung fibroblasts. After 24 h of treatment with 10 μg/ml AS, 100 μg/ml CS, and 100 μg/ml CB, immunocytochemistry was performed for COX-2 protein. Both CS and AS induce COX-2 expression as seen by the deep red cytoplasmic staining, whereas CB, a relatively inert particle, does not. Pictures were taken with a ×40 objective.

Silica induces COX-2 gene expression in primary human lung fibroblasts.

To determine whether silica directly induces COX gene expression, we treated primary human lung fibroblasts with amorphous and crystalline silica at concentrations of 10 and 100 μg/ml, respectively, and RNA was harvested and analyzed by RPA. COX-1 mRNA is expressed constitutively in untreated fibroblasts and was unchanged following treatment with either crystalline or amorphous silica (Fig. 3A). COX-2 differs from COX-1 in that it is highly inducible (35). There is no expression of COX-2 mRNA in the untreated human lung fibroblasts (Fig. 3B). However, expression is clearly induced following 6 h of treatment with either form of silica. In agreement with the Western blot results, 10 μg/ml of amorphous silica were at least as potent a stimulus as 100 μg/ml of crystalline silica.

Fig. 3.

AS and CS induce COX-2 but not COX-1 mRNA in human lung fibroblasts. Total RNA was extracted from fibroblasts after 6-h exposure to selected agents. The RNA was then analyzed by RNase protection assay (RPA). Lane 1, undigested probe; lane 2, digested probe; lane 3, untreated fibroblasts; lane 4, fibroblasts treated with 100 μg/ml CS; lane 5, fibroblasts treated with 10 μg/ml AS. A: neither form of silica induces COX-1 gene expression. B: both CS and AS induce COX-2, with AS at least 10 times more potent than CS.

Fibroblast PGE2 production in response to silica is specific and mediated by COX-2.

The conversion of arachidonic acid to PGH2 by the COX enzymes is a key rate-limiting step in the production of PGE2, an important immunomodulatory prostaglandin. Production of PGE2 by lung fibroblasts was measured in culture supernatants following stimulation with both forms of silica. Stimulation of fibroblasts with either type of silica resulted in significant production of PGE2 (Fig. 4). This increase was both dose (Fig. 4) and time dependent (data not shown). Amorphous silica was approximately seven times more potent than crystalline silica, as stimulation with 100 μg/ml of crystalline silica resulted in PGE2 production (4,710 pg/ml) intermediate between 10 and 25 μg/ml amorphous silica (4,200 and 6,500 pg/ml, respectively). Carbon black did not stimulate PGE2 production over background (data not shown).

Fig. 4.

PGE2 production is induced in a dose-dependent fashion by silica. A and B: fibroblasts were exposed to CS or AS at the indicated concentrations; the cell supernatants were harvested after 48 h and analyzed for PGE2 using an enzyme immunoassay (EIA). The increases in PGE2 production are statistically significant (*P < 0.01, **P < 0.001). Data represent means ± SE of 6 samples per condition. Note the different scales on the y-axis.

To confirm that the induction of PGE2 seen following treatment with silica was dependent on increased expression of COX-2, fibroblasts were cocultured with silica and either a specific COX-2 inhibitor (SC-58125) or a nonspecific COX inhibitor (indomethacin). Each of these inhibitors resulted in complete inhibition of PGE2 production following exposure to amorphous or crystalline silica (Fig. 5). Indomethacin is a potent, nonspecific inhibitor of both COX-1 and COX-2 and, as such, would be expected to completely inhibit the production of PGE2 by inhibiting conversion of arachidonic acid to the precursor PGH2. SC-58125 is highly selective for COX-2. The fact that both indomethacin and SC-58125 are equally effective in inhibiting silica-induced PGE2 production demonstrates that this increase is dependent on the increased expression of COX-2 induced by silica.

Fig. 5.

Silica-induced PGE2 production is dependent on increased COX-2 expression. Suspensions of CS and AS and either a nonselective COX inhibitor, indomethacin (20 μM), or a selective COX-2 inhibitor, SC-58125 (5 μM), were applied to fibroblasts in culture. IL-1β (10 ng/ml), a cytokine that is known to stimulate COX-2, is used as a positive control. Cell supernatants were harvested after 48 h, and PGE2 levels were measured using a specific EIA. Both indomethacin and SC-58125 independently attenuated the increase in PGE2 seen with CS (*P < 0.005) and AS (**P < 0.001). Data represent means ± SE of 6 samples per condition.

One route by which silica could induce PGE2 is through IL-1β, a cytokine that has proinflammatory and profibrotic properties (18). To determine whether silica induced IL-1β expression in lung fibroblasts, we tested cell supernatants for IL-1β production by a specific ELISA. The levels of IL-1β detected in each sample were negligible, and no increase was seen with either crystalline or amorphous silica (data not shown). This indicates that increased PGE2 production is not related to increased IL-1β expression.

Amorphous and crystalline silica induce mPGES.

In addition to the COX enzymes, production of PGE2 is dependent on specific PGESs to convert PGH2 to PGE2. Two isoforms of PGES have been identified, a microsomal form (mPGES) that is induced by inflammatory mediators and a cytosolic form (cPGES) that is constitutively expressed and not inducible (36). Little is known of the regulation of these specific prostaglandin synthases in the lung. To investigate whether upregulation of PGES contributes to increased production of PGE2, we performed Western blot analysis of fibroblasts exposed to crystalline silica (100 μg/ml) and amorphous silica (10 μg/ml). Both amorphous and crystalline silica induced expression of mPGES, and this expression was persistent, lasting up to 72 h after treatment (Fig. 6). Silica treatment did not induce expression of cPGES (data not shown).

Fig. 6.

CS and AS induce microsomal prostaglandin E synthase (mPGES) in primary human lung fibroblasts. Normal pulmonary fibroblasts in culture were exposed to 100 μg/ml CS, 10 μg/ml AS, or IL-1β (10 ng/ml) as a positive control. Cell lysates were harvested at 24, 48, and 72 h, and Western blot analysis for mPGES was performed.

Amorphous and crystalline silica induce PGF but not PGD2.

To examine the specificity of silica-induced induction of various prostaglandins, we examined production of PGF and PGD2 in the lung fibroblast supernatants. Both amorphous and crystalline silica stimulated production of PGF in a dose-dependent fashion (Fig. 7). The levels measured were 10-fold lower than those of PGE2 in the same supernatants. Amorphous silica was ∼10-fold more potent at stimulating the release of PGF, consistent with the results for COX-2 and PGE2. In contrast PGD2 production was negligible in untreated fibroblasts and not induced by crystalline or amorphous silica (data not shown).

Fig. 7.

A and B: PGF production is stimulated by CS and AS. Fibroblasts were exposed to indicated doses of CS and AS. The cell supernatants were harvested after 48 h and analyzed for PGF using EIA. The increase in PGF production is statistically significant (**P < 0.01).

The finding that amorphous silica induces COX-2, PGE2, and PGF at 7- to 10-fold lower concentrations than crystalline silica may help to explain why amorphous and crystalline silica have different effects in vivo. To determine whether these silicas have the capacity to differentially regulate other important proinflammatory and profibrotic mediators, we assayed supernatants from silica-treated fibroblasts for IL-6, IL-8, MCP-1, and TGF-β. Neither crystalline nor amorphous silica induced the expression of IL-6, MCP-1, or TGF-β in lung fibroblasts (data not shown). However, amorphous silica strongly induced the production of IL-8 (700 ± 166 pg/ml with 10 μg/ml amorphous silica compared with 45 ± 20 pg/ml with 100 μg/ml crystalline silica, P < 0.01). This is an interesting finding as IL-8 is a potent neutrophil chemoattractant that is capable of synergizing with PGE2 to further promote neutrophil chemotaxis.


Inhalation of both crystalline and amorphous silica results in pulmonary inflammation. However, only chronic exposure to the crystalline form leads to silicosis, i.e., fibrosis. The mechanisms by which these particles induce pathology remain unclear, but little or no attention has been focused on the role of the lung fibroblast, especially primary strains from human lungs. On the basis of our data and that of others, the fibroblast is now considered a key effector in lung inflammation and fibrosis, capable of producing inflammatory and fibrotic mediators (14, 34, 40).

The data presented here demonstrate that, remarkably, both crystalline and amorphous silica directly activate human primary pulmonary fibroblasts to produce COX-2, mPGES, and PGE2, independently of production of inflammatory mediators from alveolar macrophages and epithelial cells. Silica induces the early inflammatory response gene COX-2 as determined by Western blot, immunocytochemistry, and RPA (Figs. 13). Production of the downstream prostaglandin PGE2 was stimulated by silica and was dependent on the induction of COX-2, as it could be inhibited by the COX inhibitors indomethacin and SC-58215 (Figs. 4 and 5). In addition to PGE2, silica stimulated lung fibroblasts to produce elevated levels of PGF, but not PGD2 (Fig. 7). This is interesting as PGD2 is rapidly converted to 15d-PGJ2, which is reported to have anti-inflammatory actions and which inhibits fibroblast migration, whereas PGF functions to promote neutrophil chemotaxis and thus has proinflammatory effects (2, 17, 20, 31). It thus appears that as well as COX-2, silica must be acting to induce some of the specific prostaglandin synthases but not others. Little is known of the regulation of these synthases in lung structural cells. We therefore examined and demonstrated induction of mPGES in primary human lung fibroblasts treated with silica (Fig. 6).

Our data consistently demonstrate that amorphous silica induces COX-2, mPGES, PGE2, and PGF more potently than crystalline silica, as ∼10-fold more crystalline silica is required to observe similar effects. This is a key finding given the fact that, although amorphous silica incites pulmonary inflammation, it does not induce fibrosis (13, 30). We speculate that this is due to its greater ability to induce PGE2. An emerging body of evidence suggests that failure of lung fibroblasts to synthesize sufficient PGE2 may be important in helping to promote the development of fibrosis. Clearly, PGE2 has antifibrotic properties in vitro as it inhibits fibroblast proliferation and collagen synthesis (3, 6, 9, 10). Furthermore, BAL PGE2 levels are lower in patients with IPF than normal volunteers (4). A number of studies have linked this to a direct failure of fibroblasts derived from patients with pulmonary fibrosis to induce COX-2 and therefore synthesize PGE2 (16, 37, 38). More recently, another study demonstrated that PGE2 can inhibit the phenotypic transition that fibroblasts undergo to myofibroblasts (19). Myofibroblasts are cells that have features intermediate between smooth muscle cells and fibroblasts and have been identified as the dominant producers of matrix elements like collagen in many fibrotic models (15, 33). Thus this is a particularly significant discovery and lends further weight to the argument that PGE2 has important antifibrotic properties.

It can be argued that prostaglandins also have proinflammatory properties and that increased production of prostaglandins would be expected to lead to increased disease. It has been reported that PLA2 knockout mice are protected from acute lung injury due to sepsis or acid aspiration (24) and are also protected from bleomycin-induced pulmonary fibrosis (23), suggesting that prostanoids have important profibrotic effects. However, it must be noted that eliminating PLA2 will eliminate production of all downstream products of the arachidonic acid pathway, including leukotrienes, thromboxanes, and prostaglandins, and this is likely to have complex pleiotropic effects. COX-2 knockout mice develop enhanced fibrosis in response to bleomycin (12, 16), whereas mice lacking the 5-lipoxygenase gene, which are leukotriene deficient, are protected from bleomycin-induced fibrosis and express elevated levels of PGE2 in lavage (27). Furthermore, fibroblasts isolated from IPF patients produce elevated levels of the profibrotic thromboxane A2 (TXA2) and reduced levels of the antifibrotic PGI2 (prostacyclin) (7). Thus it appears that leukotrienes and TXA2 have profibrotic properties, whereas PGE2 and prostacyclin have antifibrotic properties. In one study, COX-2-deficient mice were protected from enhanced fibrosis by a compensatory upregulation of COX-1 that restored normal PGE2 production (12). Although prostaglandins can have both pro- and anti-inflammatory effects, the results discussed above support the hypothesis that PGE2 has antifibrotic effects and provide the context for our hypothesis that crystalline, but not amorphous, silica causes fibrosis in part because amorphous but not crystalline silica is a potent inducer of PGE2 in human lung fibroblasts.

The differential ability of amorphous and crystalline silica to induce fibrosis probably involves both the physical and inflammatory properties of the silicas. Amorphous silica is more soluble and is more easily cleared from the lungs. Although increased IL-8 production leads to greater acute inflammation (13), we hypothesize that increased production of PGE2 prevents this transient inflammatory response from developing into fibrosis. Because crystalline silica particles are far larger and hence less easily cleared from the lung, the epithelial damage they cause is more persistent, whereas macrophages that ingest crystalline silica are activated and damaged (29). This persistent irritation, combined with insufficient production of PGE2, could allow the establishment of a profibrogenic milieu (16). It should be noted that prominent inflammation does not precede tissue scarring in IPF (11, 26). Perhaps an initial prominent inflammatory response is actually protective against the future development of ongoing scarring and tissue remodeling.

In summary, we demonstrated that both crystalline and amorphous silica are both capable of directly activating human primary pulmonary fibroblasts to upregulate COX-2 and mPGES and to produce elevated levels of PGE2 and PGF. Of note, amorphous silica leads to far more intense induction of this pathway. This may help explain the differential abilities of crystalline and amorphous silica to generate a fibrogenic response. These data therefore further our understanding of the potential pathogenic mechanisms of silica-induced pulmonary inflammation and fibrosis. Furthermore, this highlights the importance of pulmonary fibroblast generated PGE2 as a protective element in fibrotic pathologies.


This work was supported by National Heart, Lung, and Blood Institute Grant HL-04492-02, The American Lung Association Grant ALA-DA-004-N, The James P. Wilmot Foundation, Phillip Morris, National Institute of Environmental Health Sciences Center Grant P30ES-01247, and Environmental Protection Agency Grant R827354.


Current address for K. M. A. O'Reilly: Dept. of Medicine and Therapeutics, Conway Inst. of Biomolecular and Biomedical Research, Univ. College Dublin, Belfield, Dublin 4, Ireland.


  • 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.


  1. 2.
  2. 3.
  3. 4.
  4. 5.
  5. 6.
  6. 7.
  7. 8.
  8. 9.
  9. 10.
  10. 11.
  11. 12.
  12. 13.
  13. 14.
  14. 15.
  15. 16.
  16. 17.
  17. 18.
  18. 19.
  19. 20.
  20. 21.
  21. 22.
  22. 23.
  23. 24.
  24. 25.
  25. 26.
  26. 27.
  27. 28.
  28. 29.
  29. 30.
  30. 31.
  31. 32.
  32. 32a.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
View Abstract