Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Pulmonary fibrosis results from severe lung injury by diverse causes (1). However, in most instances, the exact mechanism by which these injuries cause fibrosis remains unknown. Previous studies have indicated that pulmonary fibrosis, as a consequence of lung injury, is characterized by migration, adhesion, and proliferation of fibroblasts in the alveolar spaces (4). The activation of fibroblasts is initiated by various cytokines generated by inflammatory cells, including alveolar macrophages (7, 8). For instance, platelet-derived growth factor (PDGF) is a potent chemoattractant and mitogen for fibroblasts, and its gene expression has been shown to be upregulated in alveolar macrophages in patients with idiopathic pulmonary fibrosis (9, 10). Endothelin-1 also promotes fibroblast proliferation and chemotaxis, and its expression is also increased in the lung of pulmonary fibrosis (11). In addition, transforming growth factor (TGF)- is a very potent cytokine in the development of pulmonary fibrosis because of its stimulatory effect to synthesize extracellular matrix protein (12). The concert of all these fibrogenic cytokines is thought to be essential in the pathogenesis of pulmonary fibrosis (13). However, cytokines/growth factors that are involved in the migration, adhesion, and proliferation of fibroblasts have not yet been identified.

Osteopontin (OPN) is an arginine-glycine-asparatic acid (RGD)-containing protein secreted by a variety of cells, including osteoclasts, activated T cells, and activated macrophages (14, 15). OPN has been known to promote adhesion and chemotaxis of different cell types, such as macrophages and vascular smooth muscle cells (SMCs), by interacting with v integrin, including v3, v1, and v5 (16). In addition, OPN is also known to upregulate proliferation of cultured human coronary artery SMCs and primary prostate epithelial cells (19). Recently, much interest has been focused on the biologic role of OPN in the pathogenesis of various diseases (22, 23). For example, OPN in cooperation with vascular endothelial cell growth factor promotes endothelial cell migration, suggesting that OPN may be involved in the angiogenesis of cancer cells (24). The potential role of OPN in wound healing and tissue injury has been also studied with great interest. Murry and coworkers (25) reported that OPN is expressed by infiltrating macrophages in experimental cardiac injury and myocardial infarction. Miyazaki and colleagues (26) reported that the OPN messenger RNA (mRNA) level is significantly increased in the lungs of transgenic mice that are expressing tumor necrosis factor (TNF)- in type II pneumocytes, leading to pulmonary alveolitis. However, little is known concerning the biologic role of OPN in pulmonary fibrosis as a consequence of severe lung injury.

These findings together with functions of OPN in cellular adhesion and chemotactic activity prompted us to examine the hypothesis that OPN produced by activated alveolar macrophages may play an important role in the development of pulmonary fibrosis. To address this hypothesis, we examined OPN expression in the lung of a mouse model of bleomycin (BLM)-induced pulmonary fibrosis, which is considered to provide a convenient model for some aspects of fibrosis developing in the setting of lung injury (27). We have also analyzed the effect of OPN on the migration, adhesion, and proliferation of fibroblasts in vitro, and discuss the biologic significance of OPN in pulmonary fibrosis.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals and Tissue Sampling

Specific pathogen-free male ICR mice 8 wk of age were purchased from Charles River Laboratories (Tokyo, Japan). Experimental animals were anesthetized with pentobarbital (50 mg/kg intraperitoneally) and were treated with bleomycin hydrochloride (Nippon Kayaku, Tokyo, Japan) by intratracheal injection (27, 28). BLM was dissolved in 100 µl saline solution and administered at a dose of 1 mg/kg body weight. The control mice were instilled intratracheally with the same volume of saline. Mice were killed on Days 0 (untreated), 3, and 14. BLM-induced responses in the lung were evaluated histologically. Samples of the lung were frozen immediately in liquid nitrogen and stored at 80°C for Northern blot and Western blot analyses. The mortality rate for intratracheal inoculation was less than 5%.

Immunohistochemical Staining

The expression of OPN in the lungs of BLM-treated mice was assessed by immunohistochemical staining using polyclonal rabbit antihuman OPN antibody (Immuno-Biological Laboratories Co., Ltd., Gunma, Japan). This antibody was raised against synthetic peptides corresponding to the carboxyl-terminal end of human OPN (29) and was confirmed to cross-react with mouse OPN specifically by immunoblot analysis. Immunohistochemical analyses were performed as previously described (30). Briefly, sections were treated with autoclave for 15 min at 120°C in 10 mM citrate buffer, pH 6.0, to retrieve the antigen. The sections were then incubated with rabbit antihuman OPN antibody diluted 1:100 for 1 h at room temperature. Specific binding was detected through avidin-biotin peroxidase complex formation with a biotin-conjugated goat antirabbit immunoglobulin (Ig) G (Vectastain ABC kit; Vector, Burlingame, CA) and diaminobendizine (Sigma, St. Louis, MO) as substrate. Staining was absent when isotype-matched immunoglobulin was used as a control. Alveolar macrophages were identified by immunoreactivity for BM8 antibody (BMA, Rheistrasse, Switzerland), which reacts specifically to murine macrophages. Immunohistochemical staining for BM8 was performed according to the manufacturer's instructions.

Cell Lines and Culture

RAW 264.7 cells, a murine macrophage cell line, were obtained from Riken Gene Bank (Tsukuba, Japan). Cells were maintained in RPMI 1640 (Nissui, Tokyo, Japan) containing 10% fetal calf serum (FCS) (GIBCO-BRL, Gaithersburg, MD). RAW 264.7 cells at 70% confluence were incubated for 24 h in the absence or presence of BLM at 0.1 or 1.0 µg/ml. Total RNAs were extracted from cells at 0, 6, 12, and 24 h after initial stimulation and subjected to Northern blot analysis. NIH3T3 cells, a murine fibroblast cell line, were obtained from American Type Culture Collection (Rockville, MD) and were grown in RPMI 1640 containing 10% calf serum (Nissui). Wi 38 cells, a human lung fibroblast, were obtained from Riken Gene Bank and were maintained in RPMI 1640 containing 10% FCS.

Western Blot Analysis

Proteins were extracted from lung tissues of BLM-treated mice. Frozen lung tissues were homogenized in lysis buffer (1% Triton X, 50 mM Tris-HCl, pH 8.0, 150 mM NaCl). Samples containing equal amounts of protein were incubated with diethylaminoethyl sepharose beads (Amersham Pharmacia Biotech, Buckinghamshire, UK) in reaction buffer (5 mM Tris-HCl, pH 7.4, 6 M urea) overnight at 4°C. Beads were washed intensively and then eluted with reducing sample buffer. Samples were separated on 10% acrylamide gels and transferred to a nitrocellulose filter by electroblotting. The filters were blocked in phosphate-buffered saline (PBS) containing 10% dry milk, washed in PBS containing 1% dry milk and 0.5% Tween-20, and then incubated with polyclonal rabbit anti-OPN antibody (Immuno-Biological Laboratories Co., Ltd.) at room temperature for 1 h. Filters were again washed and then incubated with horseradish-peroxidase-conjugated antirabbit antibody (Amersham Pharmacia Biotech) for 1 h. Filters were then washed in Tris-buffered saline with Tween (150 mM NaCl, 10 mM Tris, pH 8, 0.05% Tween-20), and specific proteins were detected using the enhanced chemiluminescence system (Amersham Pharmacia Biotech).

Northern Blot Analysis

The full-length murine OPN complementary DNA (cDNA) was amplified from RAW 264.7 cells by using sense primer (5'-GCCTGGATCCTCCCGGTGAAAGTGACTGAT-3') containing the BamHI restriction site and the cDNA sequence coding the first seven amino acids of mature OPN, and antisense primer (5'-GTTAGAATTCCTGCTTAATCCTCACTAACA-3') containing the EcoRI restriction site and the cDNA sequence of the six amino acids starting from the 19 amino acids of C-terminal from the stop codon. The oligonucleotides used in this study were synthesized based on the published murine OPN cDNA sequence (31). Nucleotide sequences of the cDNA were verified with a DNA sequencer (Perkin Elmer, Foster City, CA). The polymerase chain reaction product was digested with BamHI and EcoRI, and subcloned into the BamHI and EcoRI sites of the pGEX-5X1 vector (Amersham Pharmacia Biotech). The template was used as a probe for Northern blots. Total RNAs were extracted from lungs and cells by the guanidium thiocyanate-phenol-chloroform extraction procedure (Tel-Test, Friendswood, TX). They were electrophoresed in 1% formaldehyde agarose gel and transferred onto a nylon membrane (Amersham Pharmacia Biotech). cDNA probes were labeled with [³²P]deoxycytidine triphosphate (Amersham Pharmacia Biotech) by the random prime method (Takara, Tokyo, Japan). Prehybridization and hybridization were carried out at 42°C overnight. Filters were washed in 2× saline sodium citrate (SSC), 0.1% sodium dodecyl sulfate (SDS) twice and in 0.1 × SSC, 0.1% SDS four times at 55°C, and exposed to film (Fuji Film, Tokyo, Japan). Filters were also hybridized with a human 28S ribosomal RNA cDNA probe as a control for loading. Autoradiography bands were quantitated by an image analyzer (Fuji Film). We confirmed that OPN mRNA was clearly detected as a single band 1.6 kb in length as previously reported (31).

Preparation of Mouse OPN-Glutathione-S-Transferase Fusion Protein

Recombinant mouse OPN-glutathione-S-transferase (GST) fusion proteins (mOPN-GST) were produced by a GST fusion protein expression system we previously reported (32). Briefly, the full-length murine OPN cDNA was amplified from RAW 264.7 cells as described previously, subcloned into the BamHI and EcoRI sites of the pGEX-5X1 vector (Amersham Pharmacia Biotech) in the same open reading frame as the GST, and then transformed in Escherichia coli. Fusion proteins were induced with isopropyl-b-D-thiogalactopyranoside (Wako, Osaka, Japan) and conjugated with glutathione agarose beads (Sigma). Purified fusion proteins were eluted with 50 mM Tris-HCl, pH 9.6, containing 5 mM reduced glutathione (Sigma), followed by dialysis with PBS. We finally confirmed that fusion proteins were detectable as a single band on SDS-polyacrylamide gel stained with Coomasie brilliant blue. Moreover, both fusion proteins and GST were confirmed to be free of endotoxin using an endotoxin kit (Sigma).

In Vitro Cell Migration Assay

Cell migration in vitro was assayed with transwell migration chambers (Becton Dickinson, Bedford, MA) with an 8-µm pore size as previously described (33). Briefly, the cell suspension (4 × 10⁵ cells/500 µl in RPMI 1640 containing 0.1% bovine serum albumin [BSA]) was added to the upper chamber, and mOPN-GST or GST was added to the lower chamber to a final concentration of 30 µg/ ml. In order to confirm suppression of cellular migration mediated by OPN, we performed additional experiments by treating cells with glycine-arginine-glycine-aspartic acid-cysteine (GRGDS) peptide (Sigma) at 100 µM or antimouse v antibody (RMV-7) (34) at 20 µg/ml. After incubation at 37°C for 6 h, the filters were fixed with 10% formalin, and stained with crystal violet. The cells on the upper surface of the filters were removed by wiping them with a cotton swab, and the cells that had migrated to the lower surface were counted under a microscope at a magnification of ×100. All assays were performed in triplicate and at least three independent experiments were performed.

In Vitro Cell Adhesion Assay

Adhesion assays were performed as previously described (32, 35). Briefly, 96-well plates were coated with mOPN-GST, GST, or BSA (Sigma) in PBS at 20 µg/ml overnight at 4°C. After three washings with PBS, wells were blocked with 1% BSA in PBS for 2 h at room temperature. Cells were resuspended in serum-free RPMI 1640 without phenol red and plated at 1 × 10⁵ cells/well. In addition, cells were pretreated with GRGDS peptide at 100 µM or RMV-7 at 20 µg/ml for adhesion analyses. Adhesion was allowed to proceed for 1 h at 37°C. The plates were inverted and centrifuged at 150 × g for 3 min; unattached cells were aspirated. The adherent cells were then placed in RPMI 1640 without phenol red containing 0.5 mg/ml thiazolyl blue (MTT) (Sigma) for 1 h at 37°C. The medium was then removed, and formazan crystals were solubilized with 50 µl dimethyl sulfoxide (Sigma). The optical density of each well was quantified at 560/655 nm wave length. The percentage of specific adhesion was determined by calculating the ratio of A560/655 of adherent cells to the A560/655 of all of the cells initially seeded. All assays were performed in triplicate and repeated at least three times.

In Vitro Cell Proliferation Assay

NIH3T3 cells were plated at 1 × 10³ cells on an 8-well culture glass slide (Lab-Tek chamber slide system; Nunc, Naperville, IL) in RPMI 1640 with 10% FCS for 48 h. After serum starvation for 24 h, cells were incubated with either GST at 10 µg/ml or varying concentrations of mOPN-GST (1, 5, and 10 µg/ml) at 37°C for 24 h in the absence or presence of human recombinant PDGF-BB (Becton Dickinson) at 1 or 10 ng/ml. Cells were also pretreated with either GRGDS peptide at 100 µM or RMV-7 antibody at 20 µg/ml for the in vitro cell proliferation assay. Subsequently, 20 µM bromodeoxyuridine (BrdU) was added to the cells. Incubation was continued for another 30 min. After three washings with PBS, cells were fixed with 70% cold ethanol, incubated in 2N HCl containing 0.2 mg/ml pepsin (Sigma) at 37°C for 20 min, and followed by treatment with 0.1 M borax (Sigma) at 4°C for 20 min. The slides were incubated in 2% H₂O₂ in methanol for 20 min to block endogenous peroxidase activity. After blocking with 10% normal rabbit serum, cells were incubated with 5 µg/ml of mouse anti-BrdU monoclonal antibody (PharMingen, San Diego, CA) for 1 h at room temperature. Specific binding was detected by peroxidase-conjugated rabbit antimouse IgG (Dako Japan, Tokyo, Japan) and diaminobendizine (Sigma) as a substrate. Nuclear staining was performed with hematoxylin. The number of cell nuclei positive for BrdU incorporation was counted in six random fields per well at ×100 magnification and divided with that of total cells. All experiments were performed in triplicate and repeated at least three times.

Administration of RMV-7 Antibody to Mice Treated with BLM

RMV-7 antibody (200 µg/mouse/d) or control isotype-matched IgG (200 µg/mouse/d) was administered intraperitoneally to mice for 11 consecutive days commencing 4 d after intratracheal BLM instillation. To assess the effect of RMV-7 antibody on pulmonary fibrosis induced by BLM, mice were killed 14 d after BLM instillation, and evaluation of fibrotic changes was performed by a numerical fibrotic scale and the measurement of hydroxyproline content in the lung as described in a subsequent paragraph.

Histologic Scoring of Pulmonary Fibrosis in Mice

For semiquantitative histologic analyses of pulmonary fibrosis induced by BLM, a numerical fibrotic scale was used according to the method described by Ashcroft and coworkers (36). Briefly, the lung of each mouse was fixed in 10% buffered formalin and stained with hematoxylin and eosin stain. More than 30 fields within each lung section were observed at a magnification of ×100, and each field was assessed individually for the severity of fibrosis and allotted a score of 0 (normal lung) to 8 (total fibrous obliteration of the field). After examination of the whole section, the mean of the scores from all fields was taken as the fibrotic score. Each specimen was scored independently by three observers. In order to prevent observer's bias, samples were coded and examined in a blind manner.

Measurement of Hydroxyproline Content in the Lung

To estimate the total amount of collagen deposition in the lung (37), the hydroxyproline content in the lungs of mice treated with BLM was measured. Briefly, after the lungs had been weighed, they were homogenized, then hydrolyzed with 6N HCl at 110°C for 16 h in a sealed glass tube. The hydroxyproline contents were determined by high performance liquid chromatography and expressed as micromole per gram.

Statistics

Statistical analysis was performed by analysis of variance (ANO-VA). All data are presented as mean ± standard deviation. Differences between means were considered statistically significant at P < 0.05 (38).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Experimental Lung Injury in Mice

We first developed a mouse model for severe lung injury leading to fibrosis by intratracheal BLM injection. A representative histologic feature of the lung in this model is shown in Figure 1. Microscopic studies of the lung in untreated mice showed normal alveolar structures (Figure 1a). Three days after BLM administration, infiltration of inflammatory cells and mild thickening of alveolar walls were observed in the lung (Figure 1b). Fourteen days after BLM instillation, alveolar walls were progressively destroyed and numerous newly accumulated fibroblasts were observed (Figure 1c), revealing the fibroproliferative change. In contrast, the control mice showed no significant change during the study after saline instillation (data not shown).

View larger version (70K): [in this window] [in a new window]   Figure 1.   Histologic appearance of the lung of BLM-treated mice. The lung tissues obtained from mice at 0 (a), 3 (b), and 14 d (c) after BLM instillation stained with hematoxylin and eosin. (Original magnification: ×100.)

Immunostaining of OPN Protein in the Lung of BLM-Treated Mice

Immunohistochemical analysis for OPN was performed on the lungs of BLM-treated mice. As revealed in Figure 2a, OPN was not detected in the lung of untreated mice. In contrast, OPN expression was evident in the lung of mice 3 d after BLM administration (Figure 2b). Moreover, there was a dramatic increase in OPN-expressing cells 14 d after BLM instillation (Figure 2c). To verify that OPN-producing cells were indeed macrophages, immunostaining using an antibody against murine macrophage (BM8) was performed. OPN-expressing cells were also stained with BM8 (data not shown), indicating that OPN was expressed principally in alveolar macrophages. As expected, control lungs during the study exhibited no significant OPN expression (data not shown).

View larger version (77K): [in this window] [in a new window]   Figure 2.   Immunohistochemical staining of OPN in the lung of BLM-treated mice. The lung tissues obtained from mice at 0 (a), 3 (b), and 14 d (c) after BLM instillation stained with polyclonal antibody against OPN. Note: OPN was expressed principally in alveolar macrophages accumulating in fibrotic area. Hematoxylin was used as a counter stain. (Original magnification: ×100.)

Induction of OPN Protein and mRNA in the Lung of BLM-Treated Mice

Western blot analysis for OPN was performed on the lungs of BLM-treated mice. As revealed in Figure 3A, OPN protein expression was induced in response to the instillation of BLM on Day 3, and was markedly increased 14 d after BLM instillation. We also examined the expression of OPN mRNA in the lungs of BLM-treated mice. In the same way, OPN mRNA was notably induced in the lung of mice with BLM instillation (Figure 3B). Neither OPN protein nor mRNA was detected in the control lungs (data not shown). Thus, in vivo studies in our model indicate that both OPN protein and mRNA were induced and notably expressed in the lung 14 d after BLM instillation.

View larger version (29K): [in this window] [in a new window]   View larger version (31K): [in this window] [in a new window]   Figure 3.   OPN expression in the lung of BLM-treated mice. (A) Proteins were extracted from lung tissues at 0 (lane 1), 3 (lane 2), and 14 d (lane 3) after BLM instillation. Equal amounts of protein were subjected to Western blot analysis using polyclonal antibody against OPN. The arrow indicates the expression of OPN, and molecular standards are shown on the left in kD. (B) Total RNAs were extracted from whole lungs at the indicated times after BLM instillation (n = 3 at Day 0, n = 5 at Day 3, n = 5 at Day 14, respectively), and 10 µg of RNAs were subjected to Northern blot analysis. Bottom panels of each set show 28S ribosomal RNA (rRNA) as control.

Induction of OPN mRNA in a Murine Macrophage Cell Line through Stimulation with BLM

Due to the results, the major source of OPN was identified as alveolar macrophages in vivo; expression of OPN mRNA was examined in RAW 264.7 cells stimulated with BLM in vitro. As revealed in Figure 4, OPN mRNA was gradually induced in RAW264.7 cells through stimulation with 1 µg/ml of BLM and increased up to 7.9-fold at 36 h after BLM stimulation compared with that of controls. Similar results were obtained from RAW 264.7 cells stimulated with 0.1 µg/ml of BLM (data not shown). Based on these results, BLM appears to stimulate macrophages to produce OPN. The finding that OPN mRNA was prominent at 36 h after BLM stimulation suggests that OPN is upregulated indirectly, for instance, via some other cytokines induced by BLM.

View larger version (39K): [in this window] [in a new window]   Figure 4.   Time course of induction of OPN mRNA in RAW 264.7 cells through stimulation with BLM. Total RNAs were isolated from RAW 264.7 cells at the indicated time intervals (h) after stimulation of cells with BLM (1 µg/ml). Total RNAs (10 µg) were subjected to blot analysis. 28S rRNA is shown as control in the lower panel.

Cell Migration of Fibroblasts Mediated by OPN

As migration and accumulation of interstitial fibroblasts into alveolar space are essential in the evolution of intra-alveolar fibrosis, we examined fibroblast migration toward OPN with the modified Boyden chamber method. As revealed in Figure 5, NIH3T3 cells significantly moved toward mOPN-GST to a greater extent than they did toward the control or GST in the lower chamber. The migration of NIH3T3 cells toward OPN was completely inhibited by the addition of GRGDS peptide or antimouse v antibody (RMV-7) to the upper chamber.

View larger version (33K): [in this window] [in a new window]   Figure 5.   Cell migration of NIH3T3 cells toward OPN. Cells were placed in the upper compartment of a Boyden-type chamber, and GST (30 µg/ ml) or mOPN-GST fusion protein (30 µg/ml) was added to the lower compartment. After 6 h of incubation, cells that migrated through the porous filter were counted at ×100 magnification. Enhanced migration of NIH3T3 cells toward mOPN-GST was abrogated by addition of either GRGDS synthetic peptide (100 µM) or RMV-7 antibody (20 µg/ml) to the upper compartment. Data are presented as the mean ± SD. *P < 0.0001 versus GST; **P < 0.0001 versus mOPN-GST.

Cell Adhesion of Fibroblasts Mediated by OPN

As attachment of migrated fibroblasts in alveolar space is also essential for intra-alveolar fibrosis, we examined fibroblast adhesion mediated by OPN. As revealed in Figure 6, it was evident that mOPN-GST propagated the adhesion of NIH3T3 cells, whereas GST did not. Adhesiveness of NIH3T3 cells to OPN was also completely inhibited by addition of GRGDS peptide or RMV- 7. These results, together with results in Figure 6, suggest that OPN expressed on alveolar macrophages may be responsible for the development of intra-alveolar fibrosis.

View larger version (28K): [in this window] [in a new window]   Figure 6.   Adhesion of NIH3T3 cells to OPN. Cells were allowed to adhere for 1 h at 37°C to 96-well plates coated with mOPN-GST, GST, or BSA at 20 µg/ml in the absence or presence of GRGDS synthetic peptide (100 µM) or RMV-7 antibody (20 µg/ml). The bound cells were quantitated using the colorimetric MTT assay. Data are presented as the mean ± SD. *P < 0.01 versus GST; **P < 0.01 versus mOPN-GST.

Cell Proliferation of Fibroblasts Mediated by OPN

To determine the effect of OPN on cell proliferation of fibroblasts, we assessed the OPN-mediated DNA synthesis of NIH3T3 cells by measuring the incorporation of BrdU. DNA synthesis was not affected by either mOPN-GST (Figure 8) or GST alone (data not shown). Interestingly, it was dramatically enhanced by mOPN-GST in the presence of 10 ng/ml of PDGF-BB in a dose-dependent manner (Figure 7). Similar results were obtained with 1 ng/ml of PDGF-BB (data not shown). Enhanced DNA synthesis was significantly suppressed by GRGDS peptide or RMV-7 antibody, indicating that the enhancement of DNA synthesis was mediated by the interaction of the GRGDS domain of OPN with v integrin on the fibroblasts. Identical findings were observed in the human fibroblast cell line Wi 38 (data not shown). These results suggest that upregulation of DNA synthesis by OPN in concert with PDGF leads to pulmonary fibrosis.

View larger version (28K): [in this window] [in a new window]   Figure 8.   Effect of either RMV-7 antibody or control IgG on BLM-induced pulmonary fibrosis. Evaluation of fibrotic changes in the lung of BLM-treated mice were performed with a numerical fibrotic scale and measurement of hydroxyproline content. (A) The grade of pulmonary fibrosis assessed by a numerical fibrotic scale. It was evaluated by the method described by Ashcroft and coworkers (36). More than 30 fields of each lung in six mice were examined. The average score of the group treated with RMV-7 antibody was significantly lower than that of the group treated with control IgG (P < 0.01). Score is expressed as mean ± SD. (B) Measurement of hydroxyproline content in the lung of BLM-treated mice. Hydroxyproline content was significantly lower in the group of RMV-7-treated mice (n = 6) in comparison with that of control IgG-treated mice (n = 6) (P < 0.05). Data are expressed as µmol/g and presented as the mean ± SD.

View larger version (34K): [in this window] [in a new window]   Figure 7.   Enhanced DNA synthesis of NIH3T3 cells by stimulation with various concentrations of mOPN-GST and PDGF-BB. After being starved in serum-free medium for 24 h, cells were left untreated or treated with mOPN-GST alone (10 µg/ml), recombinant PDGF-BB alone (10 ng/ml), or PDGF-BB and various concentrations of mOPN-GST (1, 5, and 10 µg/ml, respectively). Cells were also treated with PDGF-BB (10 ng/ml) and mOPN-GST (10 µg/ml) in the presence of the GRGDS peptide (100 µM) or RMV-7 antibody (20 µg/ml). BrdU was incorporated to the cells, and mouse monoclonal antibody against BrdU was used for immunostaining for cell nuclei to assess DNA synthesis of cells. Positive stained cells for BrdU were counted in six random fields per well at ×100 magnification and divided with total number of cells. Data are presented as the mean ± SD. *P < 0.01 versus PDGF-BB; **P < 0.05 versus PDGF-BB + mOPN-GST, 10 µg/ml.

Effect of RMV-7 Antibody on BLM-Induced Pulmonary Fibrosis In Vivo

BLM-treated mice were given either control IgG or RMV-7 antibody, which interferes with OPN-mediated fibroblast migration, adhesion, and proliferation. As shown in Figure 8A, semiquantitative histologic analysis revealed that treatment with RMV-7 antibody, but not with control IgG, significantly attenuated pulmonary fibrosis. Furthermore, we measured hydroxyproline content in the lung to assess pulmonary fibrosis. As shown in Figure 8B, the hydroxyproline content of the lungs of mice treated with RMV-7 antibody was significantly lower in comparison with the group treated with control IgG. Together with in vitro studies, these in vivo results suggest that the interaction of OPN produced by macrophages and v integrin on fibroblasts may play a crucial role in the pathogenesis of pulmonary fibrosis.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

OPN was first identified as a bone matrix protein with adhesive function because of its integrin binding activity (39). However, OPN is now known to be expressed in various tissues and different cell types, and is considered a multifunctional protein with cytokine/chemokine-like properties (15, 40, 41). OPN is upregulated under not only physiologic conditions but various pathologic conditions, and the implication of OPN in a wide variety of diseases has been investigated. In this study, we revealed that (1) OPN was strongly expressed principally in alveolar macrophages accumulating in fibrotic areas in the lung of BLM-treated mice, (2) the development of the fibrotic process was associated with an increase in the expression of OPN in the lung, (3) OPN strongly mediated fibroblast migration, adhesion, and proliferation in vitro, and (4) in vivo administration of RMV-7 antibody, which interferes with OPN binding to fibroblasts, attenuated BLM-induced pulmonary fibrosis. These findings imply that OPN plays a key role in the development of pulmonary fibrosis. To our knowledge, our study may be the first report to reveal that OPN functions as a fibrogenic cytokine that promotes migration, adhesion, and proliferation of fibroblasts in pulmonary fibrosis.

We first examined what actually upregulates OPN expression in the lung of BLM-treated mice. In rodents, the administration of BLM has been shown to directly stimulate alveolar macrophages to secrete various cytokines and free radicals, resulting in pulmonary fibrosis (42). As expected, stimulation of the murine macrophage cell line RAW 264.7 with BLM in vitro resulted in upregulation of OPN mRNA expression (Figure 4). It is unlikely that OPN is directly upregulated by BLM because OPN mRNA was induced gradually, and strong or substantial expression was attained at 36 h after BLM stimulation. OPN expression is known to be induced by various cytokines, such as TGF-, PDGF, and TNF- (45). Moreover, we have recently reported that OPN is directly induced by nitric oxide produced by activated macrophages (48). Thus, the upregulation of OPN mRNA by BLM appears to be mediated by various cytokines and/or free radicals that are produced by macrophages in response to BLM.

On the basis of previous reports, not only migration and proliferation but also adhesion of fibroblasts are essential for the initial process of pulmonary fibrosis (4, 6). OPN is strongly expressed in Langhans giant cells in tuberculosis, which are composed of fused epithelioid cells, suggesting its capability of attachment of cells (49). Previous researchers have identified fibrinogenic growth factors such as PDGF and insulin-like growth factor-1 that promote migration and proliferation of fibroblasts (7, 8). However, cytokines/growth factors that can mediate fibroblast adhesion as well as migration and proliferation have not yet been reported. In this study, we revealed the enhanced expression of OPN in pulmonary fibrosis in vivo and demonstrated that OPN mediates fibroblast migration, adhesion, and proliferation in vitro, suggesting the crucial role of OPN in the pathogenesis of pulmonary fibrosis.

We have shown that OPN in concert with PDGF-BB cooperatively promoted DNA synthesis of NIH3T3 cells (Figure 7). The exact mechanism by which OPN enhances DNA synthesis is not fully understood. However, our results that OPN alone could not upregulate the DNA synthesis of NIH3T3 cells and that the enhancement of DNA synthesis was significantly attenuated by either GRGDS peptide or RMV-7 antibody suggest that both PDGF- and v integrin-mediated signals may be responsible for the promotion of OPN-mediated DNA synthesis of fibroblasts.

Besides v integrin, CD44 is also reported to be an OPN ligand (50). However, Smith and associates (51) reported that there was no interaction between CD44 and OPN. Moreover, the lung cancer cell line transfected with the CD44 gene, which does not originally express any of the v integrins, never bound to OPN (data not shown). In this study, we have shown that the effects of OPN on in vitro migration, adhesion, and proliferation of fibroblasts were almost completely abrogated by the anti-v integrin antibody (RMV-7) to the same extent as RGD peptides. These results suggested that the majority of OPN binding to fibroblasts is mediated by the v integrins, including v3, v1, and v5, and prompted us to use RMV-7 antibody, which can inhibit OPN binding to all v-containing integrins, to eliminate in vivo OPN function in the development of BLM-induced fibrosis. In this study, we have shown that in vivo administration of RMV-7 antibody but not control IgG significantly improved BLM-induced pulmonary fibrosis in mice (Figure 8). This in vivo result strengthens our hypothesis that the OPN-v integrin interaction may be implicated in the pathogenesis of pulmonary fibrosis. Further studies employing OPN null mutant mice need to confirm our hypothesis.

The migration and proliferation of fibroblasts followed by extracellular matrix accumulation are also crucial in the pathogenesis of pulmonary fibrosis (4). It is known that OPN interacts with collagen, fibronectin, and proteoglycan (52). Liaw and coworkers (52) reported matrix organization and collagen fibrinogenesis after skin incisions are altered in the OPN null mice, whereas they remain intact in control mice. It has been shown that OPN binds directly to collagen type I, II, III, IV, and V, and forms a stable complex with fibronectin (53). These results suggest that OPN plays an important role not only in fibrinogenesis but also in the synthesis and/or turnover of matrix components.

Conclusively, our study revealed that OPN was significantly expressed in alveolar macrophages in fibrotic areas in the lung of BLM-treated mice. OPN upregulates cell migration, adhesion, and proliferation of fibroblasts in vitro. Inhibition of the interaction of OPN with v integrin by administration of RMV-7 antibody improved pulmonary fibrosis in BLM-treated mice. On the basis of these findings, we believe that our in vitro and in vivo studies provide valuable insight into the interesting role of OPN in the pathologic processes of pulmonary fibrosis and that they may be of great value in the future for the treatment of pulmonary fibrosis, but further studies are necessary.