INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Silicosis is an environmentally acquired interstitial lung disease caused by the inhalation of crystalline silicon dioxide (1), and is a major disabling lung disease throughout the world (2). Silicosis is characterized by fibrotic lesions in the lung that can lead to progressive lung impairment and eventually to death.

Current concepts suggest that the interaction between silica and alveolar macrophages (AM), and subsequent activation of AM are initial steps in the pathogenesis of the fibrotic lesions of silicosis (3). Once activated, AM phagocytose the silica particles but are unable to degrade these particles, contributing to lysis of the AM (4). The silica so released, is again ingested by AM, repeating the cycle of cell activation, injury, and death. This chronic activation of AM is thought to be a trigger for chronic inflammation and eventually for fibrosis. Upon activation by silica, AM increase their free radical production (5). Silica alone can increase free radical production through an interaction of reactive groups on the surface of silica with reactive oxygen intermediates (ROI) such as hydrogen peroxide to form hydroxy radicals (6). Iron increases the reactivity of silica with ROIs (7). Studies have also suggested that increases in levels of adhesive proteins, such as fibronectin (8) and intracellular adhesion molecule-1 (ICAM-1) (9), may play a role in the pathogenesis of silicosis.

Vitronectin is a 75,000-D adhesive glycoprotein synthesized mainly by the liver, and is largely found in the blood (10). Vitronectin binds to glass or silicon dioxide with high affinity (11), mediates cell adhesion (12), is involved in the complement system (13), helps modulate the thrombin-antithrombin III interaction in coagulation (14), and stabilizes plasminogen activator inhibitor (15). Vitronectin binds to crocidolite asbestos and facilitates internalization of the fiber into mesothelial cells (16). A recent study found elevated vitronectin levels in the blood of mice in response to endotoxin administration (16), suggesting that vitronectin may act as an acute-phase response protein.

Vitronectin is a normal constituent of the lung (16), but its function in the lung is largely unexplored. It has been speculated that vitronectin may play a role in pathogenesis of various lung disorders. For example, vitronectin is present in increased concentrations in patients with sarcoidosis (17) and hypersensitivity pneumonitis (18). Also, vitronectin interacts with potential respiratory pathogens such as streptococci, Staphylococcus aureus, Escherichia coli (19), Pneumocystis carinii (20), and Candida albicans (21). However, the role of vitronectin in the pathogenesis of lung disease is poorly understood.

The present study was initiated to examine a possible role for vitronectin in protecting AM from silica exposure. The study indicates that instillation of silica into rat lung results in an increased level of vitronectin in bronchoalveolar lavage fluid (BALF). A possible role for vitronectin in the response to the airway delivery of silica quartz is supported by the following three results: (1) vitronectin became bound in vitro and in vivo to silica particles both; (2) vitronectin significantly reduced silica toxicity to AM as determined with a ⁵¹Cr release assay; and (3) vitronectin reduced free radical generation associated with silica exposure through the two mechanisms of reducing free radical production by silica in a cell-free system and reducing silica-induced oxidant production by AM. These data suggest that increased vitronectin levels may protect the lung from silica exposure by reducing the toxicity associated with oxidant generation in the lung.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Isolation of AM and Cell-Free BALF

AM were isolated from BALF (22) obtained from Sprague-Dawley rats (Harlan Sprague-Dawley, Indianapolis, IN). The rats were rendered unconscious by placement in an ethyl ether chamber, and were then killed with an intraperitoneal injection of 0.5 ml of a solution containing 3.9 mg/ml sodium pentobarbital and 0.5 mg/ml sodium phenytoin (Schering-Plough Animal Health Corp., Kenilworth, NJ). The trachea was exposed and an 18-gauge angiocath was inserted. BALF was obtained by instilling five 10-ml aliquots of Hanks' balanced salt solution (HBSS) without calcium or magnesium (GIBCO BRL, Gaithersburg, MD) plus 0.6 mM ethylene diamine tetraacetic acid (EDTA), penicillin (100 U/ml), and streptomycin (100 µg/ml). The lavage fluid was centrifuged (1,200 × g for 5 min) and the cell pellet was resuspended in lysing solution (11 mM KHCO₃ and 152 mM NH₄Cl) to remove red blood cells. For Western blot analysis, BALF was obtained with 15 ml of HBSS and the cell-free BALF was saved for vitronectin quantification. A small aliquot from the resuspended AM was saved, cytocentrifuged, and stained to obtain cell differential counts. AM represented 98% of the cells in the differential count. AM were centrifuged (1,200 × g for 5 min), enumerated with a hemocytometer, and resuspended in RPMI-1640 (BioWhittaker, Walkersville, MD).

Purification of Vitronectin

Bovine vitronectin was purified by adapting of a previously reported procedure (23). Briefly, bovine blood was drained into 2.0 L of anticoagulant buffer (176 mM glucose, 90 mM sodium citrate, 16 mM citric acid, 16 mM sodium phosphate [monobasic], and 5 mM EDTA). Red blood cells were removed by centrifugation (4,420 × g for 10 min at 4° C). Bovine plasma was clotted in glassware by adding 30 ml of 1.0 M CaCl₂. The clotted plasma was centrifuged (12,000 × g for 15 min at 4° C), and the serum was saved. Phenylmethylsulfonyl fluoride (PMSF) and EDTA were added to the serum to achieve final concentrations of 4.0 mM and 5.0 mM, respectively. The serum mixture was applied to a Sepharose 4B precolumn (2.5 × 24 cm) (Pharmacia, Piscataway, NJ) connected in series to a heparin Sepharose column. A heparin column (2.5 × 24 cm) was first washed with 10 mM Na₂HPO₄ (pH 7.7), 5.0 mM EDTA, and 2.0 M NaCl (500 ml), and was then equilibrated in 10 mM Na₂HPO₄ (pH 7.7), 5.0 mM EDTA, 0.13 M NaCl, and 0.1 mM PMSF. The flow-through fraction was collected, and urea was added to achieve a final concentration of 8.0 M. The heparin Sepharose column was washed with 10 mM Na₂HPO₄ (pH 7.7), 5.0 mM EDTA, 2.0 M NaCl, 8.0 M urea, and 10 mM -mercaptoethanol (500 ml), and was then subjected to equilibration in 10 mM Na₂HPO₄ (pH 7.7), 5.0 mM EDTA, and 8.0 M urea (Buffer A). The denatured flow-through fraction was then reapplied to the heparin Sepharose column. The column was washed with Buffer A plus 0.13 M NaCl (500 ml), and vitronectin was eluted with Buffer A plus 0.50 M NaCl. The A₂₈₀-positive fractions were pooled and dialyzed three times in 4.0 L each of 50 mM Tris (pH 7.0) and 0.15 M NaCl. The purity of vitronectin was verified by sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (24) and by Western blot analysis (25), using a rabbit polyclonal antibody to bovine vitronectin (GIBCO BRL). The vitronectin concentration was determined with the bioinchoninic acid protein assay (Pierce, Rockford IL).

Production of Monoclonal Antibodies against Vitronectin

Monoclonal antibodies (mAb) were produced according to the method of Kohler and Milstein (26) with necessary modifications. Briefly, immune spleen cells were obtained from BALB/c mice immunized with purified bovine vitronectin. The spleen cells were mixed and fused with an 8-azaguanine-resistant myeloma cell line (SP2/0 Ag.14; American Type Culture Collection, Rockville, MD) in the presence of 1.0 ml of 50% polyethylene glycol for 1.0 min. After the fusing agent was removed, the cells were resuspended in RPMI-1640 containing hypoxyanthine, aminopterin, and thymidine (HAT); 10% fetal calf serum; glutamine (300 µg/ml); gentamicin (4 µg/ml); and nystatin (25 IU/ml). The cells were distributed into 24-well tissue culture plates (Fisher Scientific Co., Pittsburgh, PA) and were maintained in the medium just described for 2 wk. After 2 wk, medium was taken from the wells containing hybrids and was assayed with an enzyme-linked immunosorbent assay (ELISA) for the presence of antibodies to vitronectin. Positive hybrids were cloned by limiting dilution into 96-well tissue culture plates containing feeder cells. The selected clones were recloned until the cloning efficiency approached 100%, to ensure monoclonality. The antibodies were then screened with the same ELISA as described earlier, with reconfirmation by Western blot analysis, to identify clones that cross-reacted with rat vitronectin. Antibodies obtained from the cell culture supernatant were used in the study.

Instillation of Silica into Rats

Silica (-quartz; Min-U-Sil 15; U.S. Silica, Berkeley Springs, WV) was heated to 250° C for 18 h to inactivate endotoxin. After cooling to 25° C, the silica was instilled transtracheally as previously described (5). Briefly, 200- to 250-g Sprague-Dawley rats were anesthetized with ketamine (Fort Dodge Laboratories, Inc., Fort Dodge, IA) (5% [wt/vol] in sterile saline). The trachea was exposed by blunt dissection. Rats were placed at a 60° angle, and 0.2 ml of silica (100 mg/ml) in sterile saline was administered transtracheally, followed by 0.5 ml of air. Sterile saline was used as a negative control. The incision was closed with surgical staples.

Western Blot Analysis of BALF from Rats Treated with Silica

BALF (10 µl) from rats treated with silica or saline was applied to a 7.5% SDS-polyacrylamide gel and electrophoresed for 3 h at 80 V. The protein was transferred to presoaked Immobilon-P membranes (Millipore, Bedford, MA) by electroblotting 18 h at 100 mA constant current in transfer buffer (192 mM glycine, 20 mM Tris base, and 20% methanol). The membrane was blocked with 10% (wt/vol) nonfat dry milk in TBST (50 mM TrisHCl, pH 7.5; 150 mM NaCl; and 0.05% Tween-20) for 1.0 h at 25° C. The membrane was washed thrice with TBST for 10 min each, and was incubated for 1.0 h at 25° C in TBST with an mAb developed in this laboratory that is specific for rat vitronectin (1:1,000 dilution). The membrane was then again washed thrice for 10 min each in TBST at 25° C, and was incubated with goat antimouse IgG-horseradish perioxidase (HRP) (1:1,000 dilution) in TBST for 1.0 h at 25° C. The membrane was developed in an enhanced chemiluminescence system (Pierce, Rockford, IL). The 65-kD and 75-kD bands from six samples were quantified by densitometric scanning with Kodak Digital Science 1D Image Analysis Software (Eastman Kodak, Rochester, NY). The vitronectin levels in the BALF samples were quantified by comparing the band intensities with a standard curve.

Immunostaining of Vitronectin-Coated Silica

To determine whether vitronectin binds silica, we incubated silica (1.0 mg/ml) in 1.0 ml of phosphate-buffered saline (PBS) plus 1.0 µM vitronectin for 18 h at 4° C. For controls, silica was incubated in PBS with and without 1.0 µM bovine serum albumin (BSA). After this treatment, the silica was twice centrifuged (12,000 × g for 5 min) and resuspended each time in PBS. After the second spin, the crystals were resuspended in 1.0 ml of PBS containing mAb to rat vitronectin, and the resuspended silica was incubated for 2.0 h at 37° C. The silica was then again twice centrifuged (12,000 × g for 5 min) and resuspended in PBS. After an additional centrifugation, the silica was resuspended in 100 µl of PBS, and a 1:1,000 dilution of fluorescein isothiocyanate (FITC)-conjugated goat antimouse antibody was added to the mixture. As an additional control, vitronectin-coated silica particles were incubated with the secondary antibody alone. The silica particles were incubated for 2.0 h at 37° C and were washed twice with 1.0 ml of PBS. After a final centrifugation, the silica particles were resuspended in 100 µl of PBS. A 10-µl aliquot of each sample was air dried on a microscope slide. Normal and fluorescent images were photographed through a Zeiss Axioskop microscope equipped for epifluorescence (Carl Zeiss, Inc., Thornwood, NY).

For in situ immunostaining, silica was instilled transtracheally into rats as previously described. Transtracheal instillation of saline into rats served as a negative control. Seven days after the instillation of silica, the rats were killed and the lungs were removed. Cryosections of the lungs were taken at a thickness of 6 µm. The lung sections were fixed in acetone at 4° C for 10 min, and were then rehydrated in PBS at 25° C for 5.0 min. After rehydration, the lung sections were incubated with 3% H₂O₂ in methanol for 1.0 h at 25° C and rinsed with PBS. The lung sections were also rinsed with PBS after each subsequent incubation. After the rinse, the lung sections were incubated with normal goat serum for 30 min at 25° C. The in situ immunostaining was done with a HistoMark Biotin Streptaridin kit (Kirkegaard & Perry Laboratories, Inc., Gaithersburg, MD). The primary antibody was a mouse mAb specific for rat vitronectin, which was taken from culture supernatant, diluted 1:500 in PBS, and incubated with lung sections for 30 min at 25° C. The lung sections were then incubated with a secondary antibody, (a biotinylated goat antimouse IgG) for 30 min at 25° C, and then with a streptavidin-HRP conjugate for 30 min at 25° C. As a last step, the lung sections were incubated with diaminobenzidine for 10 min. The lung sections were photographed with both light and polarized light microscopy.

⁵¹Cr Release Assay as a Measure of AM Toxicity

The ⁵¹Cr release assay was done essentially as described by Martin and Kachel (27). Briefly, AM were isolated, cultured at 2.5 × 10⁵ cells/well with RPMI-1640 in a 96-well plate, and allowed to adhere for 4.0 h at 37° C in 5% CO₂. Cells were washed twice with 200 µl of RPMI-1640 and labeled with 0.5 µCi ⁵¹Cr diluted in 200 µl RPMI-1640 per well. AM were incubated with the ⁵¹Cr for 18 h at 37° C in 5% CO₂. Following the incubation, AM were washed twice with 200 µl each of RPMI-1640 to remove unincorporated ⁵¹Cr. Concurrent with ⁵¹Cr-labeling of the AM, silica (1.0 mg/ml) was incubated in RPMI-1640 alone or with 1.0 µM vitronectin or 1.0 µM BSA for 18 h at 4° C. The silica preparations at 10 µg/ml in 100 µl of RPMI-1640 were added to the ⁵¹Cr-labeled AM, centrifuged (1,000 × g for 5 min), and incubated for 4.0 h at 37° C with 5% CO₂. Following the incubation, a 50-µl aliquot was removed from each well and labeled "A." Ten percent Triton X-100 (50 µl) was added to each well and aspirated multiple times to lyse the cells. Another 50-µl aliquot was taken from each well and labeled "B." The aliquots were counted in a gamma counter (Beckman Instruments, Inc., Fullerton, CA). ⁵¹Cr release was determined as 2A/(A + 2B). To determine concentration dependence, silica was treated with 0, 0.1, 0.5, 1.0, and 2.0 µM vitronectin, and the cytotoxic index was calculated as ([% experimental release  % control release]/(100  % control release]) × 100.

Silica-Induced Free Radical Production

The generation of ·OH radicals by silica in a cell-free system was measured with a thiobarbituric acid (TBA) assay according to the method of Schapira and colleagues (5). Silica (1.0 mg/ml) was incubated in 1.0 ml PBS plus 1.0 µM vitronectin for 18 h at 4° C. For controls, silica was incubated in PBS with and without 1.0 µM BSA. Briefly, 1.0 mg of the treated silica was centrifuged (1,200 × g for 10 min), and the pellet was resuspended in 1.0 ml of PBS plus calcium containing 1.0 mM H₂O₂, 1.0 mM deoxyribose, and 1.0 mM ascorbate. The mixture was then incubated at 37° C for 1.0 h with constant agitation and centrifuged (1,200 × g for 10 min), and a 1.0-ml aliquot was saved from each sample. Equal volumes of 1.0% (wt/vol) TBA and 2.8% (wt/vol) trichloroacetic acid (TCA) were added to the 1.0-ml aliquot, and this mixture was then incubated at 100° C for 10 min. The reaction was allowed to cool to 25° C and the absorbance at 532 nm was recorded. For the kinetic assays, the generation of hydroxy radical was done as previously described but with the addition of 0.5 mM FeCl₃. Aliquots (1.0 ml) were taken at 20-min intervals, the silica was pelleted in a microfuge tube (1,200 × g for 10 min), and 0.9 ml of the supernatant was saved for the TBA assay. After the addition of equal amounts of 1.0% (wt/vol) TBA and 2.8 % (wt/vol) TCA, the aliquots were immediately placed in a boiling water bath for 10 min. The reaction was allowed to cool to 25° C and the absorbance at 532 nm was recorded.

Respiratory Burst in AM

To quantitate the respiratory burst of AM, a chemiluminescence assay was performed according to the method of Forgue and coworkers (28). Briefly, AM (5.0 × 10⁵) were placed in sterile polystyrene luminometer cuvettes (Wallace Inc., Gaithersburg, MD) with 1.0 ml of medium 199. Silica (1.0 mg/ml) was incubated in medium 199 plus 1,000 U/ml of polymyxin B with and without 1.0 µM vitronectin for 18 h at 4° C. As a control, silica was incubated in the same manner with 1.0 µM BSA. After incubation, the silica was centrifuged (12,000 × g for 15 min) and the pellet was resuspended in fresh medium 199. The resuspended silica was added to the AM to achieve a final concentration of 0.1 mg/ml, and this preparation was incubated for 30 min at 37° C with 5% CO₂. Luminol (Sigma Chemical Co.) was added to a final concentration of 60 µM, and the resulting preparation was placed in a BioOrbit 1251 luminometer (LKB Wallac, Turku, Finland). The signal in millivolts versus time was integrated for 8 min, with data expressed as mVs/5.0 × 10⁵ AM.

Statistics

The vitronectin levels in the BALF samples were compared through the use of Student's t test. In the other experiments, each sample was run in triplicate and all experiments were repeated three times. Data were analyzed through one-way analysis of variance (ANOVA), using SigmaStat version 2.0 software (Jandel Scientific, San Rafael, CA). Chemiluminescence experiments were repeated four times and the data were analyzed thorugh a general linear models procedure with repeated measures. Dunnett's adjustment was used for multiple comparisons in which untreated silica was used as a control. These calculations were performed with the SAS statistical package (SAS Institute, Cary, NC). Error bars in figures represent the SEM. Significance was defined as p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Vitronectin Levels in BALF from Rats Exposed to Silica

Western blot analysis indicated an increase in vitronectin levels in BALF from rats treated with silica (Figure 1A). Densitometric analysis of the blot (Figure 1B) showed that the vitronectin levels of the controls and the silica-exposed rats increased from 37.5 ± 11.2 µg/ml (mean ± SEM) to 151.8 ± 24.1 µg/ml, respectively. This represents a roughly fourfold increase in vitronectin levels. As previously reported (4), silica treatment increased the protein concentration in the BALF from control values of 335.2 ± 28.3 µg/ml to values of 509 ± 60.6 µg/ml (Figure 1C), which represents a 1.5-fold increase. These data suggest that 65% of the protein increase in the BALF was due to the increase in vitronectin.

View larger version (36K): [in this window] [in a new window]   View larger version (9K): [in this window] [in a new window]   View larger version (12K): [in this window] [in a new window]   Figure 1.   Vitronectin levels were increased in BALF from rats treated with silica. (A) Vitronectin levels were determined through Western blot analysis. Rats were given 40 mg silica in saline by transtracheal administration. Rats were killed after 21 d, and 15 ml of BALF was obtained from each rat. Equal volumes of BALF (20.8 µl) were loaded and subjected to electrophoresis on a 7.5% SDS-polyacrylamide gel. The protein was transferred to a membrane for Western blots analysis. (A) An mAb specific for rat vitronectin was used as a primary antibody. The Western blots of the 65 kD and 75 kD bands are shown. (B) Vitronectin levels were determined by densitometric scanning (n = 6; *p < 0.05). (C) Total protein levels in BALF from rats treated with and without silica was determined with the bicinchoninic acid assay (n = 6; *p < 0.05). These results indicate that vitronectin levels are increased along with other protein during silica exposure.

Immunostaining of Silica

Immunostaining with an mAb for rat vitronectin demonstrated that vitronectin became bound to silica particles both in vitro and in vivo (Figures 2A through 2C). The immunostaining of silica in vitro revealed variable coating of amorphous silica particles with vitronectin. Apparently, a subpopulation of intensively stained silica crystals existed (Figure 2A). Negative controls, consisting of silica alone, BSA-treated silica, and vitronectin-treated silica treated only with the secondary antibody, an FITC-conjugated goat antimouse antibody, demonstrated no fluorescence (data not shown). This finding is consistent with the high binding affinity of vitronectin for glass or silicon dioxide (11).

View larger version (78K): [in this window] [in a new window]   View larger version (80K): [in this window] [in a new window]   View larger version (114K): [in this window] [in a new window]   Figure 2.   Vitronectin binds to silica in vitro and in vivo. (A) In vitro. Silica was incubated for 18 h at 4° C in PBS containing 1.0 µM vitronectin. Vitronectin binding to silica was detected with an mAb specific for rat vitronectin, followed by a goat antimouse antibody conjugated to FITC. The pattern was detected by fluorescence microscopy. Controls (not shown), consisting of silica alone, silica incubated with BSA, and silica treated with vitronectin and incubated only with the secondary fluorescent antibody in the absence of the primary antibody, were uniformly negative. (B and C) In vivo. Rats were instilled with silica as previously described. The rats were killed 1 wk after instillation. The lungs were removed, sectioned, and prepared for in situ immunostaining with an mAb specific for rat vitronectin. The immunoperoxidase-stained section in (B) reveals a crystal-like material positively stained for vitronectin, and in (C), polarized light microscopy identifies the material in the same section as crystalline. The results indicate that vitronectin binds to silica both in vitro and in vivo.

In situ immunostaining of lung sections from rats treated with silica showed occasional brown-stained particles in the alveolar spaces. An example of such a brown-stained particle is shown in Figure 2B. The brown stain was the result of the HRP-diaminobenzidine reaction, indicating antibody recognition of vitronectin on the surface of the silica crystal. Polarized light microscopy confirmed that the particles were crystalline silica (Figure 2C). The negative control showed no vitronectin staining in the alveolar spaces (data not shown).

Effect of Vitronectin on Silica-Induced Toxicity in AM

Vitronectin reduced silica-induced AM toxicity as determined by the ⁵¹Cr release assay (Figure 3A). Incubation of AM with untreated silica (10 µg/ml) resulted in an increase in ⁵¹Cr release from 13.1 ± 0.3% to 23.8 ± 1.6% (p < 0.05). In contrast, incubation of AM with the same concentration of silica that had been pretreated with 1.0 µM vitronectin resulted in a decrease in ⁵¹Cr release from 23.8 ± 1.6% to 14.9 ± 1.1% (p < 0.05 versus untreated silica). Further, incubation of AM with BSA-treated silica produced a ⁵¹Cr release of 22.1 ± 2.4% (p > 0.5 versus untreated silica). The effect of vitronectin on silica-induced toxicity to AM was concentration-dependent and expressible as a cytotoxic index (Figure 3B). Untreated silica resulted in a cytotoxic index in AM of 12.3 ± 2.8%. Treatment of silica with increasing concentrations of vitronectin greatly reduced the cytotoxic index in AM, which leveled off with 1.0 µM vitronectin at 2.1 ± 2.2%, with no further reduction with 2.0 µM vitronectin. Consequently, all remaining experiments were performed by pretreating silica (1.0 mg/ml) with 1.0 µM vitronectin.

View larger version (14K): [in this window] [in a new window]   View larger version (14K): [in this window] [in a new window]   Figure 3.   Vitronectin reduced silica toxicity to AM as determined by 51Cr release. (A) Silica (1.0 mg/ml) was incubated in RPMI-1640 with 1.0 µM vitronectin. For controls, silica was incubated with saline or 1.0 µM BSA for 18 h at 4° C. The treated silica was added at a final concentration of 10 µg/ml to a 96-well plate containing 2.5 × 105 51Cr-labeled AM. Silica treated with 1.0 uM vitronectin significantly reduced 51Cr release from AM (n = 3; *p < 0.05). (B) Silica (1.0 mg/ml) was incubated with increasing concentrations of vitronectin (0.0 to 2.0 µM) for 18 h at 4° C. The vitronectin-treated silica was added at a final concentration of 10 µg/ml to a 96-well plate containing 2.5 × 105 51Cr-labeled AM (n = 3; *p < 0.05). The data indicate that the protective effect of vitronectin is concentration-dependent.

Effect of Vitronectin on Silica-Induced ·OH Radical Production in a Cell-Free System

Vitronectin reduced ·OH radical production by silica in a cell-free system. Free radical production as evidenced by TBA- reactive products was expressed as A₅₃₂ (Figure 4A). Untreated silica resulted in an A₅₃₂ of 0.116 ± 0.016 OD units, whereas silica treated with 1.0 µM of vitronectin resulted in an A₅₃₂ of 0.052 ± 0.004 OD units (p < 0.05). No significant difference in ·OH radical production occurred when silica was treated with 1.0 µM BSA as compared with untreated silica.

View larger version (14K): [in this window] [in a new window]   View larger version (15K): [in this window] [in a new window]   Figure 4.   Vitronectin reduced ·OH production by silica in a cell-free system. (A) Silica (1.0 mg/ml) was incubated in PBS with 1.0 µM vitronectin or with 1.0 µM BSA as a control for 18 h at 4° C. Free radical production was measured through TBA-reactive products at an absorbance of 532 nm. The data indicate that vitronectin reduced the ability of silica to generate free radicals (n = 3; *p < 0.05). (B) The TBA assay was repeated with three aliquots of each sample at different time intervals (0 to 120 min) (n = 3; *p < 0.05). The data indicate that vitronectin, but not BSA, slowed the production of free radicals by silica in the presence of 0.5 mM Fe3+.

Iron is known to increase the toxicity of silica through the Fenton reaction (7). We performed a kinetic assay with the TBA assay in the presence of 0.5 mM FeCl₃ (Figure 4B). The change in A₅₃₂ with respect to time was recorded, and the slope corresponding to the linear portion of the curve was calculated to determine the rate of reaction. The rate of reaction as calculated from the TBA assay for untreated silica was 0.0973 ± 0.0014 OD U/min, which was reduced for vitronectin-treated silica to 0.0649 ± 0.0014 OD U/min (p < 0.05). BSA-treated silica yielded a reaction rate of 0.0895 ± 0.0047 OD U/min, which was not significantly different from the rate with untreated silica (p > 0.05). These results indicate that vitronectin significantly slows the rate of ·OH radical production in a cell-free system.

Effect of Vitronectin on Silica-Induced Respiratory Burst in AM

Vitronectin reduced the silica-induced respiratory burst by AM as determined with a chemiluminescence assay (Figure 5). AM incubated with silica pretreated with 1.0 µM vitronectin showed a decrease in respiratory burst intensity from 257.8 ± 54.9 (mVs/5.0 × 10⁵ AM) with untreated silica to 121.9 ± 42.9 (mVs/5.0 × 10⁵ AM; p < 0.05) with vitronectin-treated silica. As a control, AM incubated with silica treated with 1.0 µM BSA showed a respiratory burst intensity of 367.3 ± 99.4 (mVs/5.0 × 10⁵ AM; p > 0.05), which was not statistically significantly different from that with untreated silica. This reduction in the respiratory burst intensity of AM may represent additional evidence for a protective role of vitronectin during silica exposure.

View larger version (16K): [in this window] [in a new window]   Figure 5.   Vitronectin reduced silica-induced free radical production by AM. Silica (1.0 mg/ml) was incubated in medium 199 plus 1,000 U/ml polymyxin B with 1.0 µM vitronectin. For controls, silica was incubated with saline or 1.0 µM BSA for 18 h at 4° C. After incubation, the silica was centrifuged (12,000 × g for 15 min) and the pellet was resuspended in fresh medium 199. The treated silica was added at a final concentration of 100 µg/ml in medium 199 containing 5.0 × 105 AM. Luminol was added (final concentration of 60 µM) to the silica-AM mixture. Chemiluminescence was measured in a BioOrbit 1251 luminometer by integrating the signal for 8.0 min (n = 4; *p < 0.05). The data indicate that vitronectin reduced the silica induced-respiratory burst in AM.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The results of this study indicate that as examined by Western blot analysis, vitronectin levels are increased in the BALF of rats treated with silica. Vitronectin binding to silica was demonstrated both in vitro and in vivo. Vitronectin reduced the toxicity of silica to AM, and reduced the ability of silica to generate free radicals as well as to inhibit the silica-mediated increase in the respiratory burst of AM. Taken together, these findings suggest that an increase in vitronectin may protect the lung from the toxic effects of silica.

An increase in protein levels in the alveolar spaces is an early event during silica exposure (29). A study by Seiffert and colleagues described vitronectin as an acute-phase response protein in various organs, including the lung, during endotoxin administration in mice (16). Our study showed an increase in the levels of vitronectin in the BALF of rats in response to silica exposure. This increase in vitronectin may have been the result of an enhanced rate of synthesis or of leakage of plasma during silica-induced lung injury. The BALF vitronectin levels in our study were higher than previously reported for human BALF (30), which may reflect species differences, obvious differences in the volume and method of collection of BALF, and differences in the method used to determine vitronectin levels. Nonetheless, BALF levels of vitronectin in silica-treated rats in our study were nearly quadrupled the values measured in control rats.

Our study indicates that vitronectin binds to silica in vivo and protects AM from silica-induced injury in vitro. These results suggest that the increase in vitronectin may be an appropriate host response to silica exposure. A recent study in our laboratory found that vitronectin on the surface of liposomes increased their uptake by AM (33). By analogy, vitronectin may also increase silica uptake. If it does, the ability of vitronectin to reduce silica-induced cytotoxicity to AM would be even more impressive.

The cytotoxic properties of silica are directly related to its available surface area. The cytotoxicity of silica is directly related to the available ionized silanol groups (SiOH SiO + H⁺), silica radicals (Si· and SiO·) (34), and metal contaminants such as Fe³⁺ on the surface of silica (7). Vitronectin binding may alter the surface properties of silica by reducing the availability of these reactive groups, thereby reducing the cytotoxicity of this compound. Current concepts suggest that negatively ionized silanol groups are responsible for cell lysis (35). Vitronectin may block these negatively ionized groups and protect AM, as demonstrated by the ⁵¹Cr release assay. Moreover, vitronectin reduces the ability of silica to generate free radicals as assessed by the cell-free TBA assay, suggesting that vitronectin may block the surface radicals on silica. Surface-bound iron enhances free radical generation by silica, probably through the Fenton reaction (7). Vitronectin reduces the production of free radicals by silica in the presence of 0.5 M Fe³⁺ as determined by a kinetic TBA assay. Taken together, these findings suggest that the protection of AM may involve vitro-nectin binding, thereby altering the surface properties of silica.

The interaction between silica particles and AM represents a critical event in the pathogenesis of silicosis (3). AM cannot digest or destroy the silica particles, and the ensuing host response results in lung injury. The reported effects of silica on AM include: (1) macrophage activation resulting in the release of oxidants (36); (2) macrophage stimulation resulting in the secretion of chemotactic factors (37); (3) the release of cytokines that modulate the host inflammatory response (38); and (4) the induction of apoptosis (39). These effects are dependent on the surface properties of silica (6), and the present study suggests that vitronectin binding alters these surface properties, resulting in a diminished respiratory burst. If the respiratory burst is excessive and overwhelms the antioxidant defenses, the resulting oxidative stress can lead to lung injury. Chemiluminescence data suggest that an additional protective role for vitronectin may involve the reduced oxidant stress resulting from reduction in the respiratory burst induced by silica.

Cell-adhesion glycoproteins other than vitronectin are also affected by silica exposure. Recent studies have found that silica is associated with an increase in ICAM-1 in mice (9), release of fibronectin by AM (8), and release of collagen by fibroblasts (3). These proteins may also serve important roles in the injury-repair response to acute silica exposure during silicosis.