Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Murine exposure to silica results in the development of lung fibrosis that is similar to the fibrotic changes observed in humans after occupational or environmental exposure to silica (1, 2). The mechanisms responsible for the development of silica-induced fibrotic lung injury are unknown but may involve tumor necrosis factor (TNF)- (1, 2). For example, mice exposed to silica upregulate their expression of the TNF messenger RNA (mRNA) and production of TNF in whole lung and in bronchoalveolar lavage cells (3). This enhanced TNF expression precedes the onset of fibroblast proliferation and subsequent collagen deposition in the lung (3). Neutralization of TNF with anti-TNF antibodies or the administration of soluble TNF receptors (TNFRs) can prevent or diminish the development of lung fibrosis resulting from silica exposure in mice (3, 5). The expression of TNF in the lung is positively correlated with the murine strain sensitivity to silica (4). C57BL/6 mice, a strain sensitive to silica, upregulate TNF expression (i.e., increase TNF mRNA) and increase TNF secretion by alveolar macrophages after exposure to silica (3, 4). C3H mice, a strain resistant to silica, do not upregulate TNF expression or increase TNF secretion after exposure to silica (4).

TNF is one of 10 known members of a family of ligands that activate a family of corresponding receptors (6). TNF binds to two receptors, the 55-kD and 75-kD, with similar affinity (8). The 55-kD receptor transduces cytotoxic and proinflammatory signals (6). Although the 75-kD receptor can also transduce these effects, it reportedly triggers thymocyte proliferation as well (6). Recently, mice have been genetically manipulated in which the genes for the two TNFRs have been deleted (11). Deletion of the 55-kD TNFR gene was found to be associated with resistance to endotoxic shock but enhanced susceptibility to infection with Listeria monocytogenes (11- 13). Deletion of the 75-kD receptor gene was reported to confer resistance to the dermatotoxic and lethal effects of TNF (13). The role of TNFR in lung fibrosis in general, and specifically in silica-induced lung fibrosis, has not yet been investigated.

In the present study we evaluated the importance of the two TNFRs in the pathogenesis of silica-induced pulmonary fibrosis. This was done by exposing to silica either wild-type (C57BL/6, BALB/c, or 129/J) mice or mice in which both TNFRs have been deleted (p55/p75/). Because it is known that exposure of mice to bleomycin also results in the development of lung fibrosis and upregulated TNF expression and secretion in the lung (5, 14, 15), we compared the effects of silica with those observed following bleomycin exposure.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals

Silica particles (< 5 mm) were a generous gift of Dr. Andrij Holian (University of Texas Health Science, Houston, TX). Silica particles were filtered and sterilized at 200°C as previously described (7). Silica suspension was made by adding sterile 0.9% NaCl (Baxter Healthcare Corp., Deerfield, IL) and sonicating immediately before intratracheal instillation. Stock solutions (5 U/ml) of bleomycin (Blenoxan; kindly donated by Bristol-Meyer-Squibb Pharmaceuticals, Princeton, NJ) were prepared immediately before use with endotoxin-free water. All other chemicals were of the highest grade commercially available.

Animals

All animal protocols were approved by the Tulane University Committee on the Use and Care of Animals (protocol no. 1984-1,2,3-04-076). Specific pathogen-free female C57BL/6, BALB/c (Charles River Laboratories, Kingston, NY), and 129/J (Jackson Laboratories, Bar Harbor, ME) mice weighing 20 to 25 g (6 to 10 wk old) were housed in pathogen-free cabinets and provided with water ad libitum. Mice genetically deficient in either p55 (p55/) or p75 (p75/) TNFRs were generated by gene targeting in embryonic stem cells (16, 17). The p55/ or p75/ mice were generated at Immunex (Seattle, WA) (16, 17). Independently derived mice lacking either p55 or p75 TNFR have been previously described (11). The p55 TNFR mutation was generated by homologous recombination using a targeting vector that replaces p55 TNFR sequences encoding amino acids 30-184 with PGKneo cassette. Mice carrying the p55 TNFR mutation were identified by genotype analyses using three primers specific for p55 (5'-GGA TTG TCA CGG TGC CGT TGAAG; 5'-TGA CAA GGA CAC GGT GTG TGGC; and 5'-TGC TGA TGG GGA TAC ATC CAT C) and a primer specific for the PGKneo cassette (5'-CCG GTC GAT GTC GAA TGT GTG).

The p75 TNFR mutation was generated by homologous recombination using a targeting vector that replaces p75 TNFR sequences encoding amino acids 3-26 with a PGKneo cassette. Mice carrying the p75 TNFR mutation were identified by genotype analyses using two primers specific for p75 TNFR (5'-AGA GCT CCA GGC ACA AGG GC; and 5'-AAC GGG CCA GAC CTC GGGT) and a primer specific for the PGKneo cassette (5'-CCG GTC GAT GTG GAA TGT GTG).

Mice lacking both p55 and p75 receptors (p55/ p75/) were generated by the appropriate crosses of singly deficient mice (16, 17). The p55/p75/ mice used throughout these studies were maintained on a random C57BL/6 × 129 hybrid background. In the present study we determined that both wild-type parental strains (C57BL/6 and 129/J) developed lung injury and fibrosis after exposure to silica or bleomycin.

Silica and Bleomycin Treatment

Animals were anesthetized with intraperitoneal tribromoethanol (Aldrich, Milwaukee, WI) and then exposed to silica or bleomycin as previously described (3, 4, 14, 15). Briefly, the trachea was exposed using a sterile technique, and 0.2 g/kg silica in 0.06 ml of 0.9% NaCl or 4 U/kg bleomycin in 0.05 of 0.9% NaCl were slowly instilled into the tracheal lumen. In preliminary studies, these doses were determined to consistently induce lung fibrosis with a low mortality (< 30%). Control mice received the same volume of sterile saline. After exposure, the skin incision was closed and the animals were allowed to recover on a warming plate. Animals were killed 28 d after silica and 14 d after bleomycin exposure because previous reports in the literature have described an enhanced expression of TNF mRNA in the lung as well as inflammation and enhanced collagen deposition for both silica and bleomycin at these time points (3, 4, 14, 15). Animals were killed with tribromoethanol and the descending aorta was severed and the thorax opened. The left lung was removed and stored at 80°C for subsequent analysis of hydroxyproline content.

DNA Probes

Complementary DNA (cDNA) templates used for experiments were as follows: The 1.101-kilobase (kb) murine TNF (pMuTNF) was obtained from American Type Culture Collection (Rockville, MD) and has been described elsewhere (18). The murine p55 and p75 TNFR cDNA were cloned as EcoRI fragments in bluescript as previously described (9). A 240-base pair (bp) SpeI-BglII fragment from p55 and a 450-bp XbaI-SalI fragment from p75 were obtained for Northern blots. The 240-bp fragment hybridized to a single 2.6-kb transcript for the p55 mRNA. In contrast, the 450-bp p75 fragment recognized a 3.6-kb and a more predominant 4.5-kb transcript (9). The murine 36B4 (a human acidic ribosomal phosphoprotein) cDNA was used for loading control as previously described (19).

Northern Blot Analysis

Tissues from silica-, bleomycin-, and saline-exposed mice (three per treatment used and murine strain) used for RNA extraction were dissected and immediately snap-frozen in liquid nitrogen. Total RNA was extracted from the lung using a cesium chloride method for Northern blot analysis (20). Briefly, frozen lungs were dissociated using a tissue tearer (Biospec Products Inc., Racine, WI) in 2 ml of 4 M guanidinium thiocyanate solution. Insoluble debris was removed by centrifugation for 10 min at 10,000 × g, and the resulting supernatant was layered onto 1 ml of cesium chloride and centrifuged at 85,000 × g for 4.5 h at 21°C using a Beckman tabletop centrifuge. RNA was separated by electrophoresis (20 mg/lane) on a 1.2% formaldehyde-agarose gel prior to transfer by capillary action to Immobilon-N transfer membrane (Millipore, Bedford, MA). Membranes were then dried for 2 h at 80°C. Membranes were hybridized overnight at 62°C with ³²P-deoxycytidine triphosphate-labeled (ICN, Irvine, CA) random-primed cDNA probes. The murine TNF, p55, p75, and 36B4 cDNAs described previously were used as templates and were derived from restriction endonuclease digestion of their respective plasmids. Membranes were probed first for TNF, then stripped and reproved for p55, for p75, and, as a loading control, for 36B4. Radiolabeled probes were generated with Ready-to-Go labeling kits (Pharmacia, Piscataway, NJ) and purified using TE Midi Select-D, G-50 spin columns (5prime-3prime, Inc., Boulder, CO). Blots were developed for 72 h using Biomax films and intensifying screens (Eastman Kodak, Rochester, NY). Signal density was quantitated with a Bio-Rad Gs-670 scanner using Molecular Analyst software (Bio-Rad, Hercules, CA). For the quantitation of the p55 mRNA values, membranes were exposed to a Fuji phosphorimager (Fujix BAS 1000) plate overnight and scanned. Quantitated analysis was determined with the use of McBAS 2.5 software (Fuji USA, Standford, CT). For each mRNA band the results were normalized to the internal control (36B4) and expressed as a fold increase between the bands for the control and the bands for the silica- or bleomycin-treated animals.

Lung Hydroxyproline Content

Lung collagen content was quantitated by measuring the total hydroxyproline content of the lung. Lung hydroxyproline concentration was determined spectrophotometrically according to the method of Kivirikko and colleagues (21). Briefly, stored left lungs were homogenized in 5% trichloroacetic acid (1:9, wt/vol) and centrifuged for 10 min at 4,000 × g. The pellet was then washed twice with distilled water and hydrolyzed for 16 h at 100°C in 6 N HCl. Hydroxyproline in the hydrolysate was assessed colorimetrically at 561 nm with p-dimethylaminobenzaldehyde. Hydroxyproline content was computed as micrograms of hydroxyproline per whole left lung. Results were expressed as the percent increase in comparison with control (saline-treated) values. Because wild-type animals demonstrate a smaller weight than TNFR knockout (Ko) mice when adjusted by age, we also analyzed the results of the hydroxyproline accumulation in the murine lung in response to silica and bleomycin by indexing these values to body weight. This index has been previously reported and is used to address specifically the difference in collagen accumulation in response to silica among murine strains with different body sizes (22). An index greater than 1 indicates that there has been an increase in lung hydroxyproline content relative to body weight (22).

Morphology

The heart was perfused with 0.9% NaCl to remove residual blood and the right lung was fixed in situ for 2 h by the intratracheal instillation of 10% neutral formalin (Sigma, St. Louis, MO), at a constant pressure of 30 cm H₂O, and preserved in fixative for 24 h. Lung tissues were then sectioned sagittally and embedded in paraffin. Sections 4 µm thick were generated and mounted onto positively charged slides (Fisher Scientific, Pittsburgh, PA). Slides were stained with hematoxylin and eosin or Masson trichrome staining for light microscopic examination. The sections were examined by two pathologists blinded to the exposure protocol. The degree of lung injury and fibrosis in 40 fields per lung at a magnification of ×100 was determined as previously described (4). The lung injury was graded as follows: Grade 0: normal tissue; Grade 1: minimal inflammatory or fibrotic changes; Grade 2: mild to moderate inflammatory and fibrotic changes (no obvious damage to the lung architecture); Grade 3: moderate inflammatory and fibrotic injury (thickening of the alveolar septae); Grade 4: moderate to severe inflammatory and fibrotic injury (formation of nodules or areas of pneumonitis that distorted the normal lung architecture, or the presence of large fibrous bands); Grade 5: severe inflammatory and fibrotic injury with total obliteration of the field by fibrotic tissue. In addition, inflammation was graded according to anatomical location, namely, peribronchial, septal, or alveolar. This scoring system was selected because it allows objective and reproducible scoring of the inflammatory and fibrotic lesion (4).

Statistics

All values are expressed as means ± SEM. Differences between murine strains were analyzed using analysis of variance with Fisher's PLSD test for pairwise comparison (Statview 4; Abacus Concept, Inc., Berkeley, CA). P < 0.05 was considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

TNF and TNFR mRNA Expression after Silica and Bleomycin Exposure

TNF mRNA expression was assessed in murine lung tissues 28 d after silica exposure. No evidence of TNF mRNA was found in saline-exposed animals (Figure 1). The intratracheal instillation of silica resulted in an enhanced expression of TNF mRNA in the lung tissues of C57BL/6 (10-fold increase), BALB/C (9-fold increase), 129/J (15-fold increase), and p55/p75/ TNFR (10-fold increase) Ko mice (Figure 1). Constitutive mRNA expression for both TNFR (p55 and p75) was found in the lung tissues from all wild-type mice exposed to saline (Figure 1). The expression of the p55 TNFR mRNA in the lung was not altered in C57BL/6 (0.4 ± 0.1-fold increase), BALB/c (0.8 ± 0.6-fold increase), or 129/J (0.4 ± 0.1-fold increase) mice after endotracheal exposure to silica (Figure 1). In contrast, silica exposure resulted in an increased expression of the p75 mRNA in the lungs of C57BL/6 (6-fold), BALB/c (4-fold), and 129/J (5.4-fold) mice when compared with saline-exposed mice (Figure 1).

View larger version (59K): [in this window] [in a new window]   Figure 1.   TNF and TNFR mRNA expression in silica-induced lung fibrosis. Northern blot analysis of TNF, p55 and p75 TNFRs, and 36B4 (loading control) mRNA expression in the lung 28 d after the intratracheal injection of saline as control (C) or silica (S). Four murine strains were exposed: C57BL/6 (B6), BALB/c, 129/J (129), and p55/p75/ TNFR Ko. Silica induced increased expression of TNF mRNA over saline-treated animals in all of the strains. Silica upregulated expression of the p75 receptor, but not the p55 receptor, in C57BL/6, BALB/c, and 129/J. Blots illustrate results obtained from one mouse of each strain exposed to control or silica. The same results were obtained from three additional sets of animals.

TNF mRNA expression was assessed in murine lung tissues 14 d after bleomycin exposure. Increased TNF mRNA expression was found in the lung tissues of C57BL/ 6 (15-fold), 129/J (10-fold), and p55/p75/ TNFR (10-fold) Ko mice but not in the lung tissues of BALB/c (0.7-fold) animals 14 d after bleomycin exposure (Figure 2). No evidence of TNF mRNA expression was found in saline-exposed animals (Figure 2). Similar to the results found after silica exposure, bleomycin exposure resulted in an increased expression of the p75 mRNA in the lung tissues of C57BL/6 (5-fold) and 129/J (5-fold) mice compared with saline-exposed mice (Figure 2). Bleomycin exposure did not alter the p55 mRNA expression in any of the lungs studied (C57BL/6 [0.5-fold increase] or 129/J [0.4-fold increase]). In contrast to the C57BL/6 and 129/J mouse strains, no evidence of increased expression for either the p55 (0.1-fold increase) or the p75 (0.8-fold increase) mRNA was observed in BALB/c mice after bleomycin exposure (Figure 2).

View larger version (61K): [in this window] [in a new window]   Figure 2.   TNF and TNFR mRNA expression in bleomycin- induced lung fibrosis. Northern blot analysis of TNF, p55 and p75 TNFRs, and 36B4 (loading control) mRNA expression 14 d after the intratracheal injection of saline as control (C) or bleomycin (B). Four murine strains were exposed: C57BL/6 (B6), BALB/c (BALB), 129/J (129), and p55/p75/ TNFR Ko. Bleomycin induced increased expression of TNF mRNA over saline-treated animals in Ko, C57BL/6, and 129/J, but not in BALB/c, mice. Bleomycin upregulated expression of the p75 receptor in C57BL/ 6 and 129/J, but not in BALB/c, mice. Bleomycin did not upregulate p55 expression in any of the murine strains. Blots illustrate results obtained from one mouse of each murine strain exposed to control or bleomycin. The same results were obtained for two sets of additional animals.

Lung Collagen Content

Lung collagen content was assessed by measuring hydroxyproline from each murine strain 28 d after silica exposure or 14 d after bleomycin exposure. As depicted in Figure 3, silica exposure resulted in a significant increase in hydroxyproline content in the lungs of C57BL/6 (80 ± 1% increase), BALB/c (73 ± 5% increase), and 129/J (39 ± 2% increase) mice compared with saline-exposed mice (P < 0.05). These murine strains demonstrated hydroxyproline indices of 2.07, 1.74, and 1.7, respectively. In p55/p75/ TNFR Ko mice, silica exposure did not result in a significant (9 ± 1%, P = NS) increase in hydroxyproline in the lungs, and these mice demonstrated a lung hydroxyproline index (LHI) of 1.08 (Figure 3).

View larger version (24K): [in this window] [in a new window]   Figure 3.   Lung hydroxyproline after silica exposure. Lung hydroxyproline was measured 28 d after a single endotracheal injection of silica. The values are expressed as percentage of values from controls. The data represent the mean ± SEM obtained from five animals for each group (saline-exposed values: C57BL/ 6 = 112 ± 8; BALB/c = 60 ± 1; 129/J = 108 ± 8; TNFR Ko = 151 ± 15 µg hydroxyproline/lung). *Strains were statistically significantly different compared with saline-exposed mice. Strains were statistically significantly different compared with TNFR Ko mice.

Figure 4 shows the changes in the lung hydroxyproline content of mice after bleomycin exposure. The intratracheal instillation of bleomycin resulted in significant (P < 0.05) increases in lung hydroxyproline in the lungs of C57BL/6 mice (58 ± 4% increase; LHI, 1.93) and 129/J mice (65 ± 0% increase; LHI, 1.97) but not in the lungs of BALB/c mice (29 ± 0% increase; LHI, 1.1). The p55/ p75/ TNFR Ko mice demonstrated a nonsignificant increase in hydroxyproline content compared with saline- exposed mice (33 ± 2% increase, P = NS) and demonstrated an LHI of 1.2 (Figure 4).

View larger version (23K): [in this window] [in a new window]   Figure 4.   Lung hydroxyproline after bleomycin exposure. Lung hydroxyproline levels were measured in the murine lung 14 d after a single endotracheal injection of bleomycin as described in MATERIALS AND METHODS. Values are expressed as percentage of values from controls. The data represent the mean ± SEM obtained from five animals for each group (saline-exposed values: C57BL/6 = 39 ± 3; BALB/c = 37 ± 5; 129/J = 34 ± 4; TNFR Ko = 61 ± 6 µg hydroxyproline/lung). *Strains were statistically significantly different compared with saline-exposed mice. Strains were statistically significantly different compared with TNFR Ko mice.

Morphology

Saline-exposed C57BL/6, BALB/c, and 129/J mice demonstrated normal lung morphology (Figure 5A). In contrast, intracheal instillation of silica resulted in inflammatory changes that were predominantly distributed to the peribronchiolar and perivascular regions of the lungs of these mice (Figures 5B to 5D). Silica-exposed mice had thickening of the alveolar septae and loss of alveolar spaces in 30% (C57BL/6 and BALB/c) and 60% (129/J) of the lung parenchyma.

View larger version (96K): [in this window] [in a new window]   View larger version (131K): [in this window] [in a new window]   View larger version (152K): [in this window] [in a new window]   View larger version (145K): [in this window] [in a new window]   View larger version (126K): [in this window] [in a new window]   Figure 5.   Silica-induced lung injury in mice. Representative low-power magnification (×200) photomicrographs of the lung obtained from four different murine strains 28 d after saline (A, lung from a C57BL/6 mouse) or silica exposure: C57BL/6 (B), 129/J (C ), BALB/c (D, ×100), and TNFR Ko (E) mice (n = 10 mice per murine strain and treatment). Note that silica causes interstitial inflammatory fibrosis, but the TNFR Ko remains essentially normal. TB = terminal bronchiole; PL = pleural surface. Bar = 20 µm.

As shown in Table 1, the lung injury scores for the silica-exposed C57BL/6, BALB/c, and 129/J mice were elevated in comparison with controls but were similar to each other (7.9 ± 0.8, 7.1 ± 0.3, and 7 ± 0.1, respectively). The lung injury score for the silica-exposed p55/p75/ TNFR Ko mice was 4.9 ± 0.3 and was lower than the other silica-exposed murine strains with intact TNF receptors (Table 1). Collagen as assessed by Masson trichrome staining was present within the inflammatory areas observed in the lungs of silica-exposed C57BL/6, BALB/c, and 129/J mice but not in saline-exposed mice (data not shown). The pleural surfaces of the silica-exposed C57BL/6 and 129/J mice were morphologically normal. Histologic evidence of lung inflammation was found in less than 5% of the lung parenchyma of the p55/p75/ TNFR Ko mice (Figure 5D).

Exposure to bleomycin resulted in the development of subpleural areas of inflammation (accumulations of macrophages and lymphocytes and fibroblast proliferation) that extended into the lung parenchyma and involved the bronchi and vasculature in both the C57BL/6 and 129/J mice (Figures 6B and 6C). The BALB/c mouse strain, known to be resistant to bleomycin-induced lung injury, showed minimal evidence of inflammation after exposure to bleomycin (data not shown). In contrast, the lung parenchyma of C57BL/6 and 129/J mice was consolidated with the loss of the normal alveolar architecture. Morphometric analysis demonstrated an involvement of 25% (C57BL/6) and 10% (129/J) of the lung parenchyma.

View larger version (94K): [in this window] [in a new window]   View larger version (122K): [in this window] [in a new window]   View larger version (116K): [in this window] [in a new window]   View larger version (111K): [in this window] [in a new window]   Figure 6.   Bleomycin-induced lung injury in mice. Representative low-power magnification (×200) photomicrographs of the lung obtained from four different murine strains 14 d after saline (A, lung from a C57BL/6 mouse) or bleomycin exposure: C57BL/6 (B), 129/J (C ), and TNFR Ko (D) mice. Note that bleomycin induces inflammation and fibrosis with predominance in the subpleural regions of the lung (n = 10 mice per murine strain and treatment). TB = terminal bronchiole; PL = pleural surface. Bar = 20 µm.

The lung injury scores for bleomycin-exposed C57BL/6 (7.6 ± 0.3) and 129/J (4 ± 0) mice were higher compared with the saline-exposed (0 in C57BL/6 and 129/J mice) or bleomycin-exposed BALB/c (2 ± 0.3) mice (Table 2). Stainable collagen was present in areas of lung injury in C57BL/6 and 129/J mice, but not in the lungs of BALB/c mice.

Compared with the results observed in C57BL/6 and 129/J mice, the lungs from p55/p75/ TNFR Ko mice demonstrated evidence of minimal peribronchiolar inflammation that involved only 5% of the total lung (Figure 6D). No evidence of parenchymal or subpleural inflammation was found in the lungs of p55/p75/ TNFR Ko mice, and these animals exhibited lower (2.6 ± 0.3) lung injury scores than did C57BL/6 (7.6 ± 0.3) and 129/J (4 ± 0) mice (Table 2). No evidence of collagen deposition was found after the trichrome staining in the lungs of the Ko mice (not shown).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The fibrogenesis observed in interstitial lung diseases is thought to be the consequence of an unremitting proinflammatory response (23, 24). According to this postulate an orderly expression of cytokines, termed the "cytokine network" (each one upregulating other cytokines as well as its own expression), perpetuates the inflammatory response in the lung (23). Cytokine networks have been characterized in different fibrotic injuries, such as radiation-induced lung fibrosis and granulomatous diseases of the lung (25, 26). TNF-mediated cytokine networks have been shown to be operative in the lung during the development of bleomycin-induced lung injury as well as in mycobacterial and schistosomal granulomatous lung diseases (23, 24, 26). In these cytokine networks, the initiation of the chronic inflammatory lesions is postulated to progress via a TNF-mediated macrophage inflammatory peptide-1 pathway, which is amplified by the expression of chemokines and is inhibited by the administration of anti-TNF antibodies (23, 24).

The mechanisms by which TNF mediates the development of lung fibrosis are not completely clear. TNF responses are mediated through interactions with the two TNFRs, p55 and p75. Little information is available regarding the behavior of these receptors during the development of pulmonary fibrosis. Piguet and coworkers demonstrated that the treatment of mice with recombinant soluble p55 TNFR ameliorates both silica- and bleomycin-induced lung fibrosis in mice (14). However, this approach has not been successful in other models of lung injury. For example, the treatment of mice infected with Pneumocystis carinii demonstrates an exacerbated course after adenoviral-mediated gene transfer of a p55 TNFR (5, 27). Agostini and associates demonstrated that lymphocytes isolated from patients with fibrotic lung diseases such as sarcoidosis and hypersensitivity pneumonitis express some of the TNFR superfamily members (e.g., CD40, CD27) (28). In the present study, we found that the normal murine lung constitutively expresses both p55 and the p75 TNFR mRNA and that exposure to either silica or bleomycin results in upregulation of the p75, but not the p55, TNFR mRNA. In the case of bleomycin, the upregulation of the p75 TNFR mRNA was found only in the lungs of those murine strains (C57BL/6 and 129/J) that exhibited an enhanced expression of TNF mRNA and subsequently developed lung injury and fibrosis.

To evaluate further a potential role for the TNFR in development of lung fibrosis, mice in which both TNFRs had been deleted (p55/p75/ mice) were exposed to silica or bleomycin. These mice were derived from a hybrid background (C57BL/6 × 129/J) that develops lung fibrosis after silica or bleomycin treatment. After exposure to silica or bleomycin, the p55/p75/ TNFR Ko mice demonstrated an enhanced expression of the mRNA that codes for TNF (Figures 1 and 2), but the mice did not develop significant inflammation or lung collagen deposition (Figures 5 and 6). These data support the concept that signaling through TNFRs is fundamental to the development of silica- and bleomycin-induced lung injury. These findings are consistent with recent experiments on the TNFR Ko mice exposed to inhaled asbestos fibers (Liu and colleagues, manuscript submitted). These mice also showed no fibroproliferative responses to asbestos exposure even though expression of TNF increased. The contribution of each receptor to the pathogenesis of fibrotic lung diseases is unknown at this time, and further clarification will require the study of the single TNFR Ko mice as well as the protein expression for these receptors.

Although we did not find evidence of upregulation of the p55 TNFR mRNA in the lungs of silica- and bleomycin-exposed animals, an important role for this receptor in the pathogenesis of fibrotic lung diseases cannot be excluded. Previous reports suggest that the expression of the gene encoding for the p55 receptor is controlled by a noninducible housekeeping promoter that does not respond to stimuli such as TNF, interleukin-1, transforming growth factor-, or the interferons (8). In contrast, the gene encoding for the p75 TNFR is more readily inducible (6). This inducible expression of the p75 TNFR is believed to contribute to the difference in biologic activity (8). In particular, this inducibility may be responsible for ligand passing, a phenomenon in which TNF molecules bind preferentially to the p75 TNFR and are subsequently transferred to the p55 receptor, allowing the ligand to exert its proinflammatory activities at lower concentrations. Also described are receptor shedding, which mediates binding and neutralization of circulating TNF, and cell proliferation (6).

As demonstrated in the present study, exposure to silica is associated with an upregulated expression of the TNF gene in the lung (Figure 1). The mechanisms by which silica upregulates TNF expression in the lung are not known. Silica is capable of altering gene transcription (29). Silica can regulate TNF transcription in myelomonocytic cells by interacting with the TNF promoter (29). In addition, silica induces the generation of reactive oxygen species both in vivo and in vitro (2, 30). These reactive oxygen species can activate transcriptional regulatory factors such as the nuclear factor B (NF-B), which is capable of binding to and activating the TNF promoter (31, 32). In support of the concept that reactive oxygen species may be an early injury event, pretreatment of rats with free-radical scavengers results in a decreased expression of TNF mRNA in alveolar macrophages after the intratracheal injection of silica (33). The mechanisms by which reactive oxygen species, produced in response to silica, activate NF-B may include the phosphorylation and subsequent degradation of IB that allows the transport of NF-B to the nucleus (31, 34). Consistent with this observation, the stabilization of IB in an alveolar macrophage cell line exposed to silica has been shown to result in inhibition of NF-B activation and a reduction of TNF expression (35). TNF is able to stimulate NF-B activity directly, creating the potential for an autocrine stimulation in which TNF regulates its own production (36).

Both silica and bleomycin can generate DNA damage and induce apoptosis (37). Bleomycin has been shown to fragment DNA in whole-lung preparations and to induce apoptosis in human and murine alveolar macrophages and epithelial cells (41, 42). We have found that the induction of apoptosis by bleomycin in murine alveolar macrophages is associated with an enhanced secretion of TNF and is a strain-specific response; that is, the magnitude of apoptosis and secretion of TNF in alveolar macrophages is greater in the bleomycin-sensitive (C57BL/6) murine strain than in the bleomycin-resistant (BALB/c) mice (Ortiz and colleagues, manuscript submitted). The mechanisms by which TNF mediates bleomycin-induced apoptosis are unknown but may involve the activation of TNFR-associated effector molecules that coordinate specific signal transduction pathways (42). Most of TNF's cytotoxic actions are mediated via the p55 (TNFR 1) through interaction of its "death domain" with the TNFR-associated death domain (TRADD) protein (8, 43, 44). TRADD association with the TNFR death domain modulates its own association to the receptor as well as the activation of NF-B (perpetuating the upregulated TNF expression) and the induction of apoptosis that contributes to TNF-mediated inflammatory responses (8, 43, 44). Whether or not TRADD and NF-B activation play a role in the development of silica and bleomycin induction of lung injury and fibrosis remains to be clarified.

In summary, we have demonstrated that the presence of TNFR is a fundamental component in the pathogenesis of silica-induced and bleomycin-induced lung fibrosis in mice. Further studies will be necessary to clarify the importance of the individual TNFRs during the process of lung injury.