Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Alveolitis and subsequent intra-alveolar fibrosis have been described as the characteristics of fibrosing pulmonary diseases, such as idiopathic pulmonary fibrosis (IPF) (1). The normal alveolar epithelium consists of two types of cells: type I and type II epithelial cells. When alveolitis occurs, the type I epithelial cells are damaged and desquamated. Subsequently, the proliferation of regenerating type II epithelial cells repairs the damaged alveolar lining (5, 6). If the regenerating type II cells fail to cover the crevices of the injured alveolar membrane, activated fibroblasts migrate into the alveolar space and produce collagens, which results in irreversible intra-alveolar fibrosis (2).

Hepatocyte growth factor (HGF), a heterodimeric peptide composed of an -subunit (35 kD) and a -subunit (64 kD), was initially identified and cloned as a potent mitogen for mature hepatocytes (7). HGF is known to act not only as a mitogen but also as a motogen or a morphogen on many kinds of epithelial cells in a paracrine fashion (10). The receptor of HGF is the c-Met proto-oncogene product, which is a membrane-spanning heterodimeric tyrosine kinase and is predominantly expressed in various types of epithelium (15, 16). The administration of recombinant HGF (rHGF) suppressed hepatic or renal damage, and promoted the regeneration and reconstruction of the normal hepatic and renal tissue structure (17). These findings suggest that HGF acts as a pleiotropic factor for various organs after injury.

In the lung, HGF promotes DNA synthesis in alveolar type II cells in vitro (21). In addition, Yaekashiwa and coworkers recently reported that a simultaneous or delayed administration of HGF equally repressed the fibrotic changes in murine lung injury induced by bleomycin (24), which suggests a possibility of HGF as a candidate for the treatment of lung fibrosis. On the other hand, in acute lung injury, a downregulation of the c-Met/HGF receptor in type II epithelial cells was observed (23). It is therefore speculated that the upregulation of c-Met expression in alveolar type II cells is important to improve the therapeutic benefit of HGF.

Several investigators reported that cytokines are widely involved in immunoinflammatory and fibrosing processes of the lung (25). Some cytokines, such as tumor necrosis factor- (TNF-), transforming growth factor- (TGF-), and platelet-derived growth factor (PDGF), are considered to be profibrotic; and others, such as interferon- (IFN-), are antifibrotic. In the present study we investigated whether these cytokines can upregulate c-Met expression in the A549 human alveolar type II cell line and in C57BL/6 mice. We demonstrate that IFN- stimulates the c-Met expression in vitro and in vivo, which suggests a novel role for IFN- in the repair process after lung injury.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Materials

The 75-cm² tissue culture flasks were purchased from Iwaki (Tokyo, Japan). Interleukin (IL)-1, IL-1, IL-4, IL-8, IL-10, granulocyte macrophage colony-stimulating factor (GM-CSF), PDGF, TNF-, TGF-₁, recombinant human IFN-, mouse IFN-, monoclonal mouse antihuman IFN- antibody were from Genzyme Corp (Cambridge, MA). Ham's F-12 medium, L-glutamin-penicillin-streptomycin, and actinomycin D (ActD) were from Sigma Chemicals (St. Louis, MO). Fetal calf serum (FCS), trypsin-ethylenediaminetetraacetic acid (EDTA), 20× standard sodium chloride (SSC), and salmon sperm DNA were from GIBCO BRL (Gaithersburg, MD). The 10× 3-(N-Morpholino)propanesulfonic acid buffer and Nonidet P-40 (NP-40) were from Cosmo Bio-MBI (Tokyo, Japan). Isogen was obtained from Nippon Gene (Tokyo, Japan). Formaldehyde, formamide, 50× Denhardt's solution, ethidium bromide, and glycerol buffer were from Wako Chemicals (Tokyo, Japan). The sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) standard board, polyvinylidene difluoride (PVDF) membrane, miniprotian II 1D cell, minitransblot module, bis-acrylamide, N,N,N',N'-tetramethylethylenediamine (TEMED), 2-mercaptoethanol, the alkaline phosphatase (AP)-labeled immunoblot kit, and goat antihuman immunoglobulin (Ig)G were obtained from Bio-Rad (Hercules, CA). Phenylmethylsulfonyl fluoride was from Boehringer Mannheim (Mannheim, Germany). Ultrapure NTP set nucleotide 5'-triphosphate was from Pharmacia (Tokyo, Japan). The A549 cells were from American Type Culture Collection (Rockville, MD). The 48-well chemotaxis chamber and 8-µm chemotactic filter were from Neuro Probe, Inc. (Gaithersburg, MD). [-³²P] deoxycytidine triphosphate (dCTP), and [-³²P] uridine triphosphate (UTP) were from Amersham (Tokyo, Japan). A polyclonal antibody to human c-Met (C-12) and a polyclonal antibody to mouse c-Met (sp-160) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Biotinylated goat antirabbit IgG, Vectastain ABC kit for alkaline phosphatase, alkaline phosphatase substrate kit, and neutral red solution were from Vector Laboratories, Inc. (Burlingame, CA).

The vector containing 4.6 kb of human c-Met complementary DNA (cDNA) was kindly provided by Dr. R. Zarnegar (University of Pittsburgh School of Medicine, Pittsburgh, PA).

Tissue Culture and Stimulation with Cytokine

A549 human type II alveolar epithelial cells were cultured in Ham's F-12 medium supplemented with 10% heat-inactivated FCS, L-glutamine (2 mM), penicillin (100 U/ml), and streptomycin (100 µg/ml). Cells were grown to confluence in complete media, and then washed with phosphate-buffered saline (PBS) and changed to serum-free medium. Fresh serum-free medium was added 12 h before the experimental period unless otherwise indicated. The concentrations of cytokines used were: IL-1 (100 pg/ml), IL-1 (100 U/ml), IL-4 (10 ng/ml), IL-8 (10 ng/ml), IL-10 (10 ng/ ml), GM-CSF (10 ng/ml), PDGF (1 ng/ml), TNF- (100 U/ ml), TGF-₁ (1 ng/ml), and IFN- (300 U/ml).

Northern Blot Analysis

Total RNA was prepared from the cells by acid guanidine thiocyanate-phenol-chloroform extraction, and quantitated by measuring absorbance at 260 nm. RNA (15 µg) was size-fractionated on 1% agarose/2.2 M formaldehyde gels and transferred to nylon membranes by capillary action. RNA was crosslinked to the nylon membrane by exposure to ultraviolet light and prehybridized in 5× SSC, 50 mM Na₂HPO₄, 10× Denhardt's solution, 2.5% dextran sulfate, 50% formamide, 0.75% SDS, and 10 µg/ml salmon-sperm DNA. The cDNA probes of c-Met were labeled with [-³²P]dCTP by the random primer method and hybridized to the immobilized RNA in 5× SSC, 20 mM Na₂HPO₄, 10× Denhardt's solution, 10% dextran sulfate, 50% formamide, 0.5% SDS, and 10 µg/ml salmon-sperm DNA at 42°C overnight. The membranes were washed at high stringency (0.1 ~ 0.2× SSC) and then exposed to film (XAR; Fuji, Tokyo, Japan) at 80°C with an intensifier screen.

Nuclei Extraction and Nuclear Run-On Assay

Nuclei from A549 cells stimulated or not stimulated with IFN- for 12 h were isolated in the same way as described earlier. Pellets of 10⁸ cells were then placed on ice, and 4 ml of NP-40 lysis buffer (10 mM Tris-HCl, 10 mM NaCl, 3 mM MgCl₂, and 0.5% NP-40) was added, followed by incubation for 10 min. After centrifugation, the nuclei precipitates were washed twice with 4 ml of NP-40 lysis buffer, resuspended in 100 µl of glycerol storage buffer (50 mM Tris-HCl, 40% glycerol, 5 mM MgCl₂, and 0.1 mM EDTA), and stored at 80°C. Digested cDNA plasmids (10 µg) were incubated in 0.2 N NaOH (50 µl) for 30 min at room temperature, added to 6× SSC, spotted onto the membrane filters, and dried under a vacuum overnight. The frozen nuclei were thawed, 100 µl of 2× reaction buffer (10 mM Tris-HCl, 5 mM MgCl₂, 300 mM KCl, and nucleotides added before use) and 100 µCi of [-³²P]UTP were added, and the mixture was incubated for 30 min at 30°C with gentle shaking. The messenger RNA (mRNA) was prepared in the same way as described earlier, and suspended in 50 µl of Tris-EDTA. The membranes were prehybridized overnight in the same prehybridization buffer used in the Northern blot analysis. The hybridization, washing, and exposure were performed in the same way.

Stability Assay

After stimulation with IFN- for 12 h, the cells were incubated with 10 µg/ml Act D with or without a change to IFN--free medium. At 40, 80, and 120 min after the addition of ActD, the cells were collected and Northern blot analyses were carried out as described earlier.

Western Blot Analysis

After stimulation with each cytokine for 24 h, A549 cells were collected and resuspended in 500 µl of saline. After 100 µl of 100% trichloroacetate acid were added, the suspensions were incubated on ice for 15 min. The samples were then centrifuged, and the pellets were collected. Next, 40 µl of 9 M urea, 2% Triton X-100, and 1% dithiothreitol were added, and the suspension was homogenated by ultrasound shock waves. A total of 10 µl of 10% lithium dodecyl sulfate, 0.02% bromophenol blue, and 10 µl of 1 M Tris were added and ultrasonic homogenization was done once more. After SDS-PAGE under reducing conditions, the proteins were transferred to a polyvinylidine disulfate membrane, and the membrane was incubated with an antibody against human c-Met, diluted at 1:100 in 0.05% Tween 20 in Tris-buffered saline. The membrane was incubated with biotinylated goat antirabbit IgG, incubated with avidin-biotin AP complex reagent, and then visualized using AP color development (Bio-Rad).

Migration Assay

A549 cell migration was examined using a Boyden chemotaxis chamber. A549 cells were grown to confluence in complete media, washed with PBS three times, and changed to serum-free medium. Fresh serum-free medium was added 12 h before the experiment. Cells were then incubated with or without IFN- for 12 h. Then 45 µl of the cell suspension (1 × 10⁶ cells/ml in Ham's F-12 medium) were placed in the upper chamber wells, and 25 µl of HGF (30 ng/ml) diluted in the same medium were placed in the lower chamber wells. The upper chamber was separated from the lower chamber by a polyvinylpyrolidone-free polycarbonate filter with 8-µm pores (Nucleopore, Costar, Cambridge, MA). The cells were incubated in the chamber for 6 h in 95% room air/5% CO₂, and then the filter was removed. After the top surface was scraped to remove nonmigrated adherent cells, the filter was stained by Diff-Quik solution. The number of migrated cells on the bottom surface of the filter per high-power field (magnification: ×400) was counted. Each assay was performed in triplicate.

Injection of Recombinant Mouse IFN-

Male C57BL/6N/Crj mice (6 wk of age) were obtained from Charles River Co. Ltd. (Kanagawa, Japan). IFN- (5,000 and 20,000 U/ml in 200 µl of physiologic saline) was administered by intraperitoneal injection. At 24 h, mice were killed and subjected to the histochemical studies. The animal protocol was approved by the Animal Ethics Board Committee of the University of Tokyo.

Histology and Immunohistochemistry

After each mouse was killed by deep anesthesia by intraperitoneal injection of pentobarbital, lungs were perfused with PBS and fixed by instilling 4% paraformaldehyde in PBS through the trachea. The lungs were then cut out and submerged in the same fixative. The tissue was embedded in paraffin, and 4-µm sections of the lungs were placed on glass slides for further analysis. For routine histologic examination, sections were stained with hematoxylin and eosin (H&E).

For immunohistochemical studies with antimouse-c-Met antibody, the tissue was deparaffinized in Xyrene and rehydrated with decreasing concentrations of ethyl alcohol. After blocking with 5% normal goat serum, the primary antibody (polyclonal rabbit antimouse-c-Met antibody) was applied at a concentration of 1.0 µg/ml and the slides were incubated at 4°C overnight. After washing with PBS, biotinylated antirabbit IgG was applied and the slides were incubated at 37°C for 30 min. After washing, avidin- biotin alkaline phosphatase complex was applied and the slides were incubated at 37°C for 30 min, followed by the addition of the substrate solution. Color development was stopped by rinsing the slides in distilled water for 5 min. The slides were counterstained with neutral red. As control for a nonspecific staining, purified rabbit IgG (1.0 µg/ml) was applied instead of the primary antibody.

Statistical Analysis

Migration data are expressed as means ± standard error of the mean (SEM) for each category examined in triplicate and were analyzed using Student's t test. When the P value was less than 0.05, the analysis was considered to be statistically significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of Cytokines on the c-Met mRNA Expression in A549 Cells

To investigate whether cytokines can upregulate the expression of c-Met mRNA, we stimulated A549 cells with cytokines for 12 h. Figure 1A shows the results of the Northern blot analysis. IL-1, IL-1, IL-4, IL-8, IL-10, GM-CSF, PDGF, TNF-, and TGF-₁ did not promote the expression of c-Met mRNA. We tried the same experiments with 0.1× and 10× concentrations of cytokines, but no effect was observed (data not shown). In contrast, the c-Met mRNA expression was strongly enhanced by the stimulation with IFN-. We checked whether these effects were specific by conducting an abrogation assay with neutralizing antibody (Figure 1B). The IFN- effect was completely inhibited by the neutralizing antibody. These results indicate that IFN- specifically upregulated the expression of c-Met mRNA.

View larger version (56K): [in this window] [in a new window]   Figure 1.   Modulation of c-Met mRNA expression in A549 cells by cytokines. Total RNA was isolated from cells as described in MATERIALS AND METHODS after treatment with cytokines for 12 h and subjected to a Northern blot analysis for c-Met mRNA detection. A photograph of the ethidium bromide-stained gel corresponding to the same blot is presented to demonstrate the RNA integrity and equal loading of RNA samples. (A) Effect of cytokines. (B) Inhibition of the effect of IFN- by anti-IFN- neutralizing antibody. IFN- was preincubated with anti-IFN- neutralizing antibody (1 ng of antibody/U of IFN-) for 1 h at 37°C. A549 cells were stimulated with IFN- for 12 h and total RNA was isolated and subjected to a Northern blot analysis.

Next, we determined the optimal concentration and stimulation time at which the c-Met mRNA reached the maximum expression (Figures 2A and 2B). IFN- upregulated the c-Met mRNA expression optimally at 300 U/ml. We then analyzed the correlation between the c-Met mRNA expression and stimulation time. Compared with the control lane, the IFN- stimulation for 12 h promoted the c-Met mRNA expression optimally.

View larger version (48K): [in this window] [in a new window]   Figure 2.   The dose-response and time-course studies. (A) A549 cells were stimulated with IFN- at the described concentrations for 12 h and submitted to a Northern blot analysis. IFN- (300 U/ ml) promoted the maximal expression of the c-Met mRNA at 300 U/ml. (B) A549 cells were stimulated with IFN- (300 U/ml) for the described periods. The expression of the c-Met mRNA was maximal when the cells were stimulated for 12 h. The left lanes of each sample show the controls (incubated for the same time without IFN-).

Effect of IFN- on the Transcription of c-met Proto-oncogene

To analyze the exact mechanism of this effect of IFN-, we examined whether IFN- upregulates the transcription of c-met proto-oncogene by conducting a nuclear run-on assay. The results are shown in Figure 3. After stimulation with IFN- for 12 h, the nuclei were extracted and the transcriptional activity was analyzed. Compared with the transcription of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene, which is constitutively expressed, the transcription of the c-met gene was clearly upregulated with IFN-. This indicates that IFN- enhances the transcription of the c-met gene.

View larger version (49K): [in this window] [in a new window]   Figure 3.   Nuclear run-on assays measuring the rate of the c-Met mRNA transcription in A549 cells stimulated or not stimulated with IFN- for 12 h. The nuclei were extracted and submitted to the assay as described in MATERIALS AND METHODS. The results were normalized using the vector plasmid and GAPDH RNA as a control.

Next, we investigated whether IFN- can affect the stabilization of c-Met mRNA. After stimulation with IFN- for 12 h, the cells were incubated with ActD (10 µg/ml) with or without a change to the IFN--free medium. ActD inhibits transcription; therefore, the effects on mRNA stabilization were observed after the addition of ActD. The reduction of the c-Met mRNA was analyzed by Northern blot analysis (Figure 4). After the addition of ActD, the intensity of the c-Met blot decreased following the time course, in both the IFN--treated and nontreated samples. No significant difference in the intensity of the decay was observed between the two groups. These results indicate that IFN- did not affect the stability of the c-Met mRNA.

View larger version (42K): [in this window] [in a new window]   Figure 4.   Stabilization study of the c-Met mRNA. A549 cells were pretreated with IFN- for 12 h, and ActD (10 µg/ml) was added. In some samples, the culture medium was changed to an IFN--free one. The cells were incubated further for 2 h with or without IFN- (300 U/ml). Total mRNA was extracted from each group of cultures at each time point and analyzed for c-Met mRNA presence by Northern blot analysis. A photograph of the ethidium bromide-stained gel corresponding to the blot is presented to demonstrate the RNA integrity and equal loading of RNA samples.

Effect of IFN- on c-Met Protein Expression

The effect of IFN- on c-Met protein expression was analyzed with a Western blot assay. A549 cells were stimulated with IFN- (300 U/ml) for 24 and 48 h, and the proteins were extracted and submitted. Figure 5 shows the result. Compared with the control lane (no stimulation), IFN- promoted c-Met protein expression. This result indicates that IFN- also upregulates c-Met expression at the protein level.

View larger version (83K): [in this window] [in a new window]   Figure 5.   Upregulation of c-Met protein expression in A549 cells by IFN-. Total protein was prepared from cells as described in MATERIALS AND METHODS after treatment with IFN- (300 U/ml) for 24 h and 48 h, and was submitted to a Western blot assay.

Effect of IFN- on HGF-Induced Chemotaxis

HGF has been reported to have a chemotactic effect on several types of tumor cell lines and alveolar epithelial cells (29). To confirm that the c-Met protein upregulated by IFN- is biologically functional, we checked the receptor function with a chemotaxis assay. As shown in Figure 6, IFN- significantly enhanced the migration activity. Next, we observed the synergic effect of IFN- (300 U/ml) plus other inflammatory cytokines, such as IL-1 (100 pg/ml), IL-1 (100 U/ml), IL-4 (10 ng/ml), IL-8 (10 ng/ml), IL-10 (10 ng/ml), GM-CSF (10 ng/ml), PDGF (1 ng/ml), TNF- (100 U/ml), and TGF-₁ (1 ng/ml). No synergic effect with IFN- on HGF-induced chemotaxis was observed (data not shown).

View larger version (44K): [in this window] [in a new window]   Figure 6.   Effect of IFN- prestimulation on the migration of A549 cells induced by HGF. A549 cells were preincubated with IFN- (300 U/ml) for 24 h and the migration activities induced by HGF (30 ng/ml) were measured by a Boyden chamber assay as described in MATERIALS AND METHODS. The results are shown as mean values ± SEM of the triplicate studies. *A significant difference was detected between the control (without HGF) and HGF-treated groups, and between the HGF-treated group and the IFN- plus HGF-treated group (P < 0.02).

Effect of IFN- on c-Met Expression In Vivo

Finally, to investigate whether IFN-'s effect on c-Met receptor expression in A549 cell is also confirmed in vivo, we administered recombinant mouse (rm)IFN- in C57BL/6 mouse by intraperitoneal injection and observed the expresion of c-Met receptor by histochemical analysis. Routine H&E study could not clarify the difference between saline-injected and IFN--injected samples (data not shown). However, immunohistochemical study with anti-c-Met receptor antibody demonstrated that ex vivo administration of IFN- (5,000 and 20,000 U) upregulated the c-Met receptor expression in vivo (Figures 7A-7C). No positive staining was obtained with control rabbit IgG (1.0 µg/ml), indicating that the positive staining was specific for c-Met receptor protein on epithelial cells (Figure 7D).

View larger version (144K): [in this window] [in a new window]   View larger version (138K): [in this window] [in a new window]   Figure 7.   Effect of IFN- on c-Met expression on alveolar epithelial cells in vivo. rmIFN- was injected intraperitoneally into a C57BL/6 mouse. At 24 h lungs were subjected to histologic study with anti-c-Met antibody (1.0 µg/ml). Original magnification: ×400. (A) Control saline-injected lung, (B) rmIFN-:5,000 U, (C ) rmIFN-:20,000 U, (D) rmIFN-:20,000 U, and stained with purified rabbit IgG (1.0 µg/ml).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Our experiments clearly show that the mRNA and protein of the c-Met proto-oncogene is upregulated in A549 human type II epithelial cells by IFN-.

The c-Met proto-oncogene was originally isolated from a chemically treated human osteosarcoma cell line, MNNG- HOS, on the basis of its ability to transform normal fibroblasts (30). The protein translated by this oncogene consisted of a truncated transmembrane tyrosine kinase receptor with a constitutively active tyrosine kinase (31). Since then, the full-length c-Met receptor has been found in a variety of tissues, predominantly of epithelial origin (32). HGF, the ligand for c-Met, is a heparin-binding polypeptide capable of exerting morphologic, mitogenic, motogenic, and metastogenic influences on a wide variety of cells, especially those of epithelial origin (36). Several studies have shown that HGF and its receptor c-Met play important roles in the cell growth and repair of many tissues (18).

In lung fibrogenic diseases such as IPF, epithelium replication is one of the important processes for tissue repair. Recent studies detected higher HGF levels in the bronchoalveolar lavage fluid and the sera of patients with IPF than in those of healthy control subjects (39, 40), indicating important roles for HGF in the fibrosing or repair process. In addition, Yaekashiwa and colleagues reported that a simultaneous or delayed administration of rHGF represses the fibrotic changes in murine lung injury induced by bleomycin (24), which shows a possibility of HGF as a new candidate for therapy. Of course, much should be further elucidated and improved for clinical application. Yanagita and associates reported the downregulation of c-Met in acute lung injury (23). We speculate that if this downregulation could be adjusted, the therapeutic effect of HGF would be more potent.

Because many reports have suggested the importance of cytokines in pulmonary fibrosing diseases (26), we investigated whether these cytokines can upregulate the expression of c-Met, using Northern blot analyses and a Western blot assay. IL-1, IL-1, IL-4, IL-8, IL-10, GM-CSF, PDGF, TNF-, TGF-₁, and IFN- were tested, and only IFN- strongly promoted the expression of c-Met mRNA and protein expression (Figures 1 and 5). These results indicate that the administration of HGF plus IFN- would be more effective than that of HGF alone for the repair of lung injury.

IFN- is a product of immune effector cells in response to many stimuli (41). The ability of IFN- to influence fibrosis directly has been a controversial subject (41). In this study we observed a novel role of IFN- in the alveolar repair process. Generally, the fibrosing pathologic process of the lung can be divided into two phases. The first is the inflammatory phase, in which inflammatory cells infiltrate the intra-alveolar spaces and secrete toxic mediators, inducing alveolar wall damage (3, 4, 28). In this phase, in addition to the suppression of inflammatory cells and toxic mediators, the rapid repair of alveolar wall and the blocking of fibroblast invasion are critical factors (6, 44). Our present results suggest that IFN- plays a role in this phase in enhancing the epithelial cell regeneration indirectly by upregulating the receptor expression of its growth factor, HGF. Saunders and colleagues have already reported that IFN- (300 U/ml) alone enhances the proliferation of A549 cells (45). A similar result was also reported in murine lung epithelial cell lines (46). Their and our results suggest that IFN- can work as an antifibrotic factor in the acute inflammatory phase of the fibrosing process. Moghul and associates demonstrated that inflammatory cytokines such as IL-1 and TNF-, which are known to be released in response to tissue injury, markedly upregulate the c-Met mRNA in several types of carcinoma cell lines (47). However, we observed no effect of these cytokines on A549 type II epithelial cells. This may indicate that there are different mechanisms for the regulation of c-Met expression among a variety of cells or tissues. TGF-₁ and PDGF are known to be strong chemotactic factors for fibroblasts. Elevations of TGF-₁ mRNA and protein after the administration of bleomycin have been reported (48). HGF expression is markedly inhibited by TGF-₁ (49). We therefore tested the possibility that these cytokines have some negative effects on c-Met expression in A549 cells, but we found no such effect.

The second phase is the fibrotic or reparative phase, in which activated fibroblasts infiltrate the alveolar space and produce collagens, resulting in irreversible intra-alveolar fibrosis (3, 4). In this phase the suppression of collagen production by activated fibroblasts is most important in blocking the disease progression. Sempowski and coworkers suggested the possibility that IFN- directly inhibits the collagen production of fibroblasts (50). Together, these findings suggest that IFN- may therefore play an important role as an antifibrotic factor in both phases, in different manners.

Encouraged by the finding of IFN-'s effect on the c-Met expression, we investigated the mechanism of this effect. The results of the nuclear run-on assay and the transcriptional inhibition by ActD suggest that IFN- acts on the transcriptional level but does not induce c-Met mRNA stability prolongation (Figures 3 and 4). IFN- has been reported to activate many kinds of gene transcriptions through the activation of the Janus kinases (JAKs) and the signal transducers and activators of transcription (STATs) (51). The JAK-STAT pathway would be included in this upregulation of c-met gene. This possibility is yet to be examined. Moghul and colleagues reported the extremely short half-life of the c-Met mRNAs, indicating the involvement of post-transcriptional mechanisms for the regulation of c-Met expression (47). c-Met mRNA has an AUUUA instability element in its 3' prime-untranslated region, and this element exists in mRNAs for several oncogenes and cytokines that are involved in the regulation of important cellular functions such as growth and differentiation (54). The mechanism of c-Met mRNA instability remains to be investigated.

Finally, we checked whether the c-Met receptor protein that was upregulated by IFN- works as a functional ligand for HGF. We examined whether IFN- can further enhance the migration activity induced by HGF. The results show that IFN- clearly enhances the HGF-induced chemotaxis (Figure 6). Taking the results obtained in this study together, it is clear that this enhancement is due to the upregulation of the c-Met receptor protein. On the other hand, other possibilities should also be considered. One possibility is that IFN- may have increased the functional receptor activity regardless of upregulation of receptor density. For example, in the case of integrin, an intracellular element, such as the Tac subunit of the IL-2 receptor, joins the cytoplasmic domain of ₁ or ₃ integrin, and thus alters the integrin affinity by this "inside-out signaling" (55). So there is a possibility that IFN- alters some intracellular molecules and thus modifies the receptor affinity of c-Met. The intracytoplasmic signaling pathways of c-Met receptor have recently been elucidated, but are not yet fully understood. So far, there has been no clear report describing such a mechanism in the HGF/c-Met receptor system. Therefore, although this possibility remains unclear and should be further investigated, it is at least certain that IFN- upregulates c-Met receptor quantitively, and that this upregulation leads to an enhancement of HGF function.

Finally, to investigate whether IFN-'s effect on c-Met receptor expression in A549 cell is also confirmed in vivo, we administered rmIFN- to C57BL/6 mice by intraperitoneal injection and observed the expression of c-Met receptor by histochemical analysis (Figure 7). Ex vivo administration of rmIFN- clearly upregulated the c-Met expression in vivo, which indicates that our result using this cell line is also confirmed in vivo.

In summary, we found that IFN- works cooperatively with HGF against the fibrosing process by various mechanisms. This finding suggests a possibility of a new therapeutic strategy for lung fibrosis. The mean survival of patients with IPF has been estimated at 3 to 6 yr, and there is still no definitely effective therapy to stop the progression of the disease. To further clarify the effect of the IFN-/HGF combination observed in our in vitro study, in vivo trials should be carried out in animal models of pulmonary fibrosis.