Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Airway mucosal injury induced by noxious agents is closely related to airway hyperresponsiveness, which is one of the most important characteristics of bronchial asthma. Replication of airway epithelial cells is a crucial step for recovery from such mucosal injury, and is likely regulated both positively and negatively by a variety of growth factors and cytokines.

Hepatocyte growth factor (HGF), also known as scatter factor, was originally identified as a potent mitogen for rat hepatocyte (1). Subsequent studies have revealed that HGF is a mesenchymally derived heterodimeric protein consisting of a 60-kD heavy chain and a 30-kD light chain linked together by disulfide bonds. HGF is now considered to be a multifunctional growth factor, which has mitogenic, motogenic, and morphogenic activities in a wide variety of organs (2, 3). The receptor for HGF is a transmembrane tyrosine kinase, which is the product of the c-met proto-oncogene (4). HGF and c-met mRNA are both highly expressed in the lung (5), and, therefore, HGF may play an important role in the growth regulation of the lung epithelial cells. In fact, it has been reported that HGF induced the proliferation of rat alveolar type II cells in vitro (6). It remains unknown to date, however, whether HGF has any effect on the proliferation of normal human bronchial epithelial cells.

Interferon- (IFN-) is a potent lymphokine produced by activated T cells, especially of Th1 type. We have demonstrated recently that IFN- inhibits proliferation of serum-stimulated human bronchial epithelial cells (7). The negative growth regulatory signals to bronchial epithelial cells are considered to be important for appropriate repair of an injured area by inhibiting excessive growth of bronchial epithelial cells.

In the present study, we investigated the effects of HGF and IFN- on the proliferation of human bronchial epithelial cells isolated from peripheral airways of adult volunteers. We found, for the first time, that HGF and IFN- exerted positive and negative effects, respectively, on human bronchial epithelial cell proliferation. We also examined the underlying mechanisms by which HGF and IFN- regulate bronchial epithelial cell proliferation, by using an immortalized human bronchial epithelial cell line, BEAS-2B, which, we found, responded to HGF in a manner similar to normal bronchial epithelial cells.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents

Recombinant human HGF was purchased from Genzyme/Techne (Tokyo, Japan).

Recombinant human IFN- was from Intergen Company (New York, NY). PD98059 was obtained from New England BioLabs, Inc. (Beverly, MA). SB203580 (8) and LY294002 (9) were from Calbiochem-Novabiochem Corp. (San Diego, CA).

A rabbit polyclonal phosphospecific anti-ERK1/2 antibody and a rabbit polyclonal anti-ERK1/2 antibody, which recognize catalytically active ERK1/2 and total ERK1/2, respectively, were from New England BioLabs. Mouse monoclonal anti-cyclin D1 and rabbit polyclonal anti-p27^kip1 antibodies were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Alkaline phosphatase-conjugated donkey anti-mouse and anti-rabbit IgG were from Jackson ImmunoResearch Laboratories, Inc. (West Grove, PA).

Cell Culture

Human peripheral bronchial epithelial cells were isolated from adult volunteers as described previously (10, 11), with some modifications, using a new type of bronchoscope (BF-2.7T; Olympus, Tokyo, Japan) and a cytology brush (BC-0.7T; Olympus). The diameter of the BF-2.7T is 2.7 mm, and the tip of the BC-0.7T has an outer diameter of 0.7 mm. BF-2.7T could be inserted through the biopsy channel of conventional bronchoscopes. Bronchial epithelial cells were collected with a cytology brush (BC-0.7T) while observing the small airways directly with a BF-2.7T fiberscope. The number of cells collected from small airways was approximately 1.0 × 10⁶, and the averaged viability was 70%. Immunohistochemical analysis showed that no less than 98% of the cells were positive for keratin and negative for vimentin, indicating that the majority of the cells were bronchial epithelial cells (11). The cells were cultured on collagen-coated tissue culture plates (Iwaki, Tokyo, Japan) at 37°C in a fully humidified atmosphere containing 5% CO₂, with a serum-free defined medium (SABM) (BioWhittaker, Walkersville, MD), which contained bovine pituitary extracts, hydrocortisone, human epidermal growth factor, epinephrine, transferrin, insulin, retinoic acid, bovine serum albuminfatty acid free, and triiodothyronine. The cells at 4-5 passages were used for the present experiments. An immortalized human bronchial epithelial cell line, BEAS-2B (12), was kindly donated by Dr. C. C. Harris (National Cancer Institute, Bethesda, MA) and was maintained as reported previously (13). BEAS-2B cells at 6-8 passages were used for the experiments.

Direct Enumeration of Epithelial Cells

Human peripheral bronchial epithelial cells were plated at 5 × 10⁴ per dish (35 mm in diameter) in SABM containing 10% fetal calf serum (FCS) in the absence or presence of human recombinant HGF (50 ng/ml), with media changed on Day 2 and on Day 4. On Day 3 and on Day 6, the cells were detached by trypsinization, and the number of cells was counted in triplicates by a standard hemocytometer. BEAS-2B cells were plated at 5 × 10⁴ per dish (35 mm in diameter) in RPMI 1640 medium (GIBCO, Grand Island, NY) supplemented with L-glutamine with or without 10% FCS (GIBCO).

Colorimetric 3-4,5-(Dimethyl-Thyazol-2-yl)-2,5-Diphenyltetrazolium Assay

The effect of HGF on proliferation of human peripheral bronchial epithelial cells and BEAS-2B cells was also evaluated by a colorimetric 3-4,5-(dimethyl-thyazol-2-yl)-2,5-diphenyltetrazolium (MTT) assay as reported previously (7, 11). In brief, the cells were cultured as described previously, except that the cells were seeded in 96-well tissue culture plates (Iwaki) at a density of 5 × 10³ cells/well in 100 µl test media, with media changed on Day 2 and on Day 4. On Day 3 and on Day 6, the cells were treated with 10 µl of the MTT labeling reagent (Chemicon International Inc., Temecula, CA) and incubated for 4 h. A dark blue formazan product of MTT was extracted by isopropylalcohol-HCl mixture, and absorbance at 570 nm was measured using an automatic ELISA reader (Bio-Rad, Hercules, CA). The data were expressed as percentages of the control value obtained from the cells cultured in the absence of HGF.

Western Blot Analysis

After subconfluence, the cells were treated with growth factor-free medium for 24 h. Growth-arrested cells were stimulated with various concentrations of HGF and specific inhibitors for different time periods. To examine the inhibitory effect of PD98059, SB203580, and LY294002, the cells were preincubated with various concentrations of each of these inhibitors for 1 h and then stimulated with 50 ng/ml HGF. To evaluate the effects of HGF on the induction of p27^kip1, the cells were cultured in RPMI 1640 medium containing 10% FCS and then stimulated with growth factors.

For the extraction of the protein, the cells were washed twice with Ca²⁺- and Mg²⁺-free Dulbecco's phosphate-buffered saline and lysed in a 2× sodium dodecylsulfate (SDS) sample buffer (125 mmol/Liter Tris-HCl [pH 6.8], 4.6% wt/vol SDS, 20% glycerol, 10% 2-mercaptoethanol). The samples were heated in a boiling-water bath for 5 min to denature the proteins fully and then centrifuged at 12,000 × g for 5 min to remove insoluble debris. Equal amounts of cellular proteins (30 µg/lane) were separated by SDS-polyacrylamide gel electrophoresis, and transferred onto Immobilon P membranes (Millipore Corp., Bedford, MA). The membranes were probed with one of the antibodies described previously, and the bands corresponding to the protein of interest were visualized with alkaline phosphatase-conjugated secondary antibodies. In some experiments, the blots were visualized with an enhanced chemiluminescence detection kit (Bio-Rad).

Statistical Analysis

Statistical significance was analyzed using analysis of variance. A probability (P) value less than 0.05 was considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

HGF-Stimulated Proliferation of Human Peripheral Airway Epithelial Cells and BEAS-2B Cells in the Presence of Serum

We first examined whether HGF has any effect on the proliferation of human bronchial epithelial cells obtained from healthy adult volunteers (Figures 1A and 1B). As we reported previously (13), the cells proliferated in the presence of serum alone (Figure 1A, closed squares). We found, in the present study, that HGF further stimulated proliferation of human bronchial epithelial cells when added to serum-containing growth media. Thus, when 5 × 10⁴ cells were plated per dish and incubated for 6 d with media changed every other day, the cells proliferated to an eventual cell number of approximately 3.5 × 10⁵ cells per dish. When human recombinant HGF (50 ng/ml) was added to the cultures together with serum (14), it induced a 1.5-fold increase in the number of cells over the control value after 6 d (Figure 1A). The same 1.5-fold increase in the cell number over the effect of serum alone was also observed when the cell proliferation was evaluated by a colorimetric MTT dye reduction assay, which reflects viable cell number (10, 15) (Figure 1B). This effect of HGF was dose-dependent, with a maximally effective concentration being 50 ng/ml (Figure 1B). HGF also induced a 1.7-fold increase in the proliferation of BEAS-2B cells, an immortalized human bronchial epithelial cell line, in a similar dose-response curve (Figures 1C and 1D).

View larger version (47K): [in this window] [in a new window]   Figure 1.   HGF-stimulated proliferation of human peripheral bronchial epithelial cells obtained from healthy volunteers (A, B) and a human bronchial epithelial cell line BEAS-2B cells (C, D) in culture. (A, C) The cells (5 × 104 per dish) were incubated in the presence (black circles) or absence (black squares) of HGF (50 ng/ml) in media containing 10% FCS, and the cell number was counted on Day 3 and on Day 6. The cell number was also evaluated in serum-free conditions (white circles, with HGF, and white squares, without HGF). *P < 0.05 versus controls (n = 3). (B, D) Assessment of cell proliferation by colorimetric MTT assay. The cells were incubated with 10% FCS and indicated concentrations of HGF in the presence of FCS for 6 d. Absorbance at 570 nm was measured and the mean + SE values were expressed as percentages of the control value (100%), which stands for the cells grown in the absence of HGF. *P < 0.05 versus controls (n = 3).

Next we examined the effect of HGF on cell proliferation of human bronchial epithelial cells and BEAS-2B cells in the absence of serum. HGF did not show any effect on cell proliferation of either type of cell when serum was absent in culture media (Figures 1A and 1C). The number of cells on Day 6 was 101% ± 9.00% and 99% ± 5% of the control value for normal human bronchial epithelial cells and BEAS-2B cells, respectively (n = 3, P > 0.05). This observation clearly indicates that costimulation with serum was necessary for mitogenic action of HGF.

HGF-Induced Activation of ERK1 and ERK2 in Human Bronchial Epithelial Cells and BEAS-2B Cells in the Absence of Serum

Shown in Figure 2A (upper panel) are the effects of HGF on the activation state of ERK1 and ERK2 in human normal bronchial epithelial cells, which are evaluated by immunoblot analysis using a phosphospecific anti-ERK1/2 antibody that recognizes phosphorylated, catalytically active forms of ERK1 and ERK2 specifically. Stimulation of serum-deprived human normal bronchial epithelial cells with HGF (50 ng/ml) induced rapid activation of ERK1 and ERK2 within 5 min, which gradually decreased thereafter to reach a plateau level that was substantially higher than the basal unstimulated level by 30 min. The activation of ERK1 and ERK2 by HGF was sustained for at least 120 min (Figure 2A). Imunoblotting with an anti-ERK1/2 antibody, which recognizes both active and inactive forms of ERK1 and ERK2, showed that an equal amount of ERK1/2 protein was loaded on each lane (Figure 2A, lower panel). It is notable that HGF stimulated ERK1/2 of human bronchial epithelial cells in the absence of serum.

View larger version (35K): [in this window] [in a new window]   Figure 2.   HGF caused phosphorylation and activation of ERK1 and ERK2 in normal human bronchial epithelial cells (A) and BEAS-2B cells (B and C). The cells were stimulated with 50 ng/ml of HGF for indicated time periods (A and B) or with various concentrations of HGF (1-100 ng/ml) for 10 min (C). The cellular proteins were separated by 10% SDS-polyacrylamide gel electrophoresis, followed by Western blot analysis, using either a phosphospecific anti-ERK1/2 antibody that specifically recognizes active forms of ERK1/2 (upper panels), or an anti-ERK1/2 antibody that reacts with both active and inactive forms of ERK1 and ERK2 (lower panels). Representative results from three different studies are shown.

Such was also the case for BEAS-2B cells (Figure 2B). The effect of HGF on ERK activation was dose-dependent (Figure 2C, upper panel) with a maximally effective concentration being approximately 50 ng/ml, consistent with the effect of HGF on proliferation (Figure 1D).

HGF Induced the Expression of Cyclin D1 but Failed to Downregulate p27^kip1 Cyclin-Dependent Kinase Inhibitor in the Absence of Serum

To evaluate the role of HGF in the regulation of cell cycle molecules involved in G1 to S phase progression, we examined whether HGF has any effect on the expression level of a G1 cyclin, cyclin D1. Serum- and growth factor-starved BEAS-2B cells were stimulated with 50 ng/ml of HGF for various time periods, and subjected to anti-cyclin D1 immunoblot analysis (Figure 3A). As reported for other cell types (14), quiescent BEAS-2B cells showed a barely detectable level of cyclin D1 protein. When the cells were stimulated with HGF in the absence of serum, it induced the upregulation of cyclin D1 protein. Thus, the cellular level of cyclin D1 increased approximately 3.8-fold over the quiescent level by 4 h and remained elevated for 8 h, returning then to the basal level by 16 h.

View larger version (38K): [in this window] [in a new window]   Figure 3.   Inhibition by PD98059 of HGF-induced ERK activation and cell proliferation in BEAS-2B cells. (A) BEAS-2B cells were preincubated with various concentrations of PD98059 (1-100 µM) for 1 h and then stimulated with 50 ng/ ml of HGF for 10 min. Representative results from three different experiments are shown. (B) BEAS-2B cells were incubated with 10% FCS and 50 ng/ml of HGF in the presence or absence of PD98059 (1-50 µM) for 6 d and then subjected to MTT assay. The data represent the means + SE of the control value in the absence of PD98059 (50 ng/ml of HGF alone), which was calculated as 100%. *P < 0.05 versus controls (n = 3). (C) PD98059 failed to inhibit HGF-induced cyclin D1 expression. Serum-deprived BEAS-2B cells were preincubated with 50 µM of PD98059 for 1 h, and then stimulated with 50 ng/ml of HGF for 8 h. Equal amounts of whole cell lysates were analyzed for the expression levels of cyclin D1 protein by immunoblotting. Representative results from three different studies are shown.

Under the same experimental conditions, we also studied the effect of HGF on the protein level of p27^kip1, which is the major cyclin-dependent kinase (CDK) inhibitor that is downregulated when cells proceed through the G1 phase to enter the S phase.

As shown in Figure 3B, serum- and growth factor-starved, quiescent BEAS-2B cells showed an elevated level of p27^{kip 1} expression. Different from cyclin D1, HGF by itself did not have any effect on the protein level of p27^kip1, which remained elevated for up to 16 h.

The MEK1-ERK and Phosphatidylinositol 3-Kinase-Pathways are Required for Cell Proliferation

It is well established (16) that PD98059, the specific inhibitor of MEK1, inhibits MEK1-mediated phosphorylation and activation of ERK1 and ERK2. Indeed, we found that PD98059 at 50 µM inhibited HGF-stimulated phosphorylation of ERK in BEAS-2B cells completely (Figure 4A). We then examined whether the MEK-ERK pathway is required for HGF- and serum-stimulated cell proliferation. The cells were cultured with 50 ng/ml of HGF in serum-containing media in the presence or absence of PD98059 (1-50 µM) for 6 d, with media changed every other day and then subjected to MTT assay. As shown in Figure 4B, we found that PD98059 inhibited HGF-induced proliferation of BEAS-2B cells potently in a dose-dependent manner similar to that for inhibition of ERK1 and ERK2 (Figure 4A). The treatment with PD98059 at 50 µM suppressed growth below the level stimulated with serum alone, suggesting that cell proliferation induced by serum alone also involved the MEK-ERK pathway. In sharp contrast to PD98059, SB203580, which is the specific inhibitor of p38 mitogen-activated protein (MAP) kinase, did not show any significant inhibition of HGF-induced ERK activation (data not shown) and cell proliferation (Table 1) for up to 10 µM. PD98059 (1-100 µM) did not show any inhibition of HGF-stimulated activation of Akt, or anisomycin-stimulated activation of p38 MAP kinase or C-Jun N-terminal kinase (JNK), indicating that PD98059 inhibited the MEK-ERK pathway selectively, and that it is not cytotoxic (data not shown).

View larger version (19K): [in this window] [in a new window]   Figure 4.   HGF-induced cyclin D1 protein expression in BEAS-2B cells but failed to downregulate the protein level of p27kip1. (A) A time course of cyclin D1 expression induced by HGF in BEAS-2B cells. BEAS-2B cells were incubated with serum- and growth factor-free medium for 24 h and then stimulated with 50 ng/ml of HGF for the indicated time periods (h) in the absence of FCS. Thirty micrograms of total cellular proteins were separated by 12.5% SDS-polyacrylamide gel electrophoresis and transferred onto nitrocellulose membrane. Western blot detection of cyclin D1 was performed by using a specific antibody. (B) BEAS-2B cells were deprived of serum for 24 h and then stimulated with 50 ng/ml of HGF for various time periods as indicated (h). Western blot detection of p27kip1 was performed as described in MATERIALS AND METHODS. Representative results from three different studies are shown.

We also examined the role of phosphatidylinositol (PI) 3-kinase in HGF-induced cell proliferation of BEAS-2B cells. We found that LY294002, the specific inhibitor of PI 3-kinase, inhibited the mitogenic action of HGF partially (30%; Table 1), without inhibiting the effect of HGF on ERK activation (data not shown). These results indicate that both the MEK-ERK- and the PI 3-kinase-pathways are necessary for proliferation of the bronchial epithelial cell line. To investigate the role of the ERK1/2- and the PI 3-kinase signaling pathways in the regulation of cyclin D1, we examined whether PD98059 or LY294002 inhibited HGF-induced cyclin D1 protein expression in BEAS-2B cells. The cells were preincubated with the inhibitors for 60 min and then stimulated with 50 ng/ml of HGF for 8 h. As shown in Figure 4C, PD98059 at 50 µM did not inhibit HGF-induced cyclin D1 protein expression to a detectable extent. Such was also the case for LY294002 at 50 µM (data not shown).

We tested whether the inhibitory effects of LY294002 and PD98059 on cell proliferation were due to the cytotoxicity. The trypan blue dye exclusion test demonstrated that more than 95% of the cells were viable at the ranges used in the experiments. Lactate dehydrogenase activities in conditioned media were not elevated, compared with those in media alone (data not shown).

IFN- Potently Inhibited Serum- and HGF-Stimulated Proliferation of Human Bronchial Epithelial Cells and BEAS-2B Cells

We have demonstrated recently that IFN- inhibited proliferation of human bronchial epithelial cells potently (7). Indeed, as shown in Figure 5, IFN- (1-50 ng/ml) inhibited serum-stimulated proliferation of normal human bronchial epithelial cells (Figure 5A) and BEAS-2B cells (Figure 5C) in a dose-dependent manner. We found that IFN- also potently inhibited the mitogenic effect of a combination of HGF and serum in normal bronchial epithelial cells (Figure 5B) as well as BEAS-2B cells (Figure 5D).

View larger version (32K): [in this window] [in a new window]   Figure 5.   IFN- inhibited HGF-induced proliferation of normal peripheral bronchial epithelial cells (A and B) and BEAS-2B cells (C and D) in a dose-dependent manner. Serum-deprived cells were incubated with HGF (50 ng/ml) with or without FCS and various concentrations of IFN- at indicated concentrations for 6 d. DNA content was assessed by MTT assay. The results are the means + SE of three determinations and are expressed as percentages of the control value obtained with HGF plus 10% serum. *P < 0.05 versus HGF plus serum group (n = 3), **P < 0.05 versus serum alone (n = 3).

IFN- Did not Inhibit HGF-Induced ERK Activation or Cyclin D1 Expression but Upregulated p27^{kip 1} CDK Inhibitor

To get insights into the mechanism by which IFN- inhibits HGF-induced cell proliferation, we first examined whether IFN- inhibited HGF-induced activation of ERK. BEAS-2B cells were preincubated with 50 ng/ml of IFN- for various time periods up to 60 min and then stimulated with HGF (50 ng/ml) for 10 min. Immunoblot analysis with phosphospecific anti-ERK1/2 antibody showed that IFN- did not inhibit HGF-induced phosphorylation and activation of ERK1 and ERK2 (Figure 6A).

View larger version (25K): [in this window] [in a new window]   Figure 6.   IFN- failed to inhibit HGF-induced ERK activation or upregulation of cyclin D1 in response to HGF. (A) Growth factor-deprived BEAS-2B cells were preincubated with 50 ng/ml of IFN- for up to 60 min as indicated, and then the cells were stimulated with 50 ng/ml of HGF for 10 min in the continued presence of IFN-. The activation states of ERK1 and ERK2 were evaluated as described in the legend to Figure 2. Representative results from three different experiments are shown. (B) Growth factor-deprived BEAS-2B cells were incubated with 50 ng/ml of HGF with or without 50 ng/ml of IFN- for 8 h. Cyclin D1 expression levels were analyzed by Western blotting as described in the legend to Figure 4.

Next we examined the effect of IFN- on HGF-induced expression of cyclin D1. BEAS-2B cells were incubated with 50 ng/ml of HGF in the presence or absence of 50 ng/ml of IFN- for various time periods. Western blot analysis showed that IFN- did not inhibit HGF-induced cyclin D1 expression either at 8 h (Figure 6B) or at 16 h (data not shown).

These findings indicate that IFN- inhibits proliferation of human bronchial epithelial cells without inhibiting the effect of HGF on either the activation of ERK or the expression of cyclin D1. At this point, we examined the possible involvement of p27^kip1 CDK inhibitor in IFN--mediated suppression of bronchial epithelial cell proliferation. IFN- (50 ng/ml) was introduced to exponentially growing BEAS-2B cells, and the p27^kip1 protein level was monitored for up to 16 h. We found that IFN- upregulated the protein level of p27^kip1 after 8 to 16 h despite continued stimulation with serum (Figure 7A). This effect of IFN- was dose-dependent, with the maximal effect obtained at 50 ng/ml (Figure 7B). Further, we found that IFN- induced upregulation of p27^kip1 CDK inhibitor in cells stimulated maximally by a combination of serum and HGF (Figure 7C). Thus, when HGF was introduced to serum-stimulated cells, the p27^kip1 level declined gradually for 16 h, whereas when IFN- was introduced together with HGF, the protein level increased.

View larger version (37K): [in this window] [in a new window]   Figure 7.   IFN- upregulated p27kip1 protein. (A) BEAS-2B cells were grown to subconfluency in RPMI 1640 containing 10% FCS. IFN- (50 ng/ml) was then introduced to the cells. They were further incubated for up to 16 h as indicated. Western blot detection of p27kip1 was performed as described in MATERIALS AND METHODS. Representative results from three different studies are shown. (B) BEAS-2B cells were placed on 35-mm tissue culture dishes at a density of 1 × 106 cells per dish with RPMI 1640 containing 10% FCS. After subconfluency, BEAS-2B cells were stimulated with various concentrations of IFN- for 16 h. (C) BEAS-2B cells were grown to a subconfluent state in RPMI 1640 medium containing 10% serum. They were further incubated with 50 ng/ml of HGF in the presence or absence of 50 ng/ml of IFN- for different time periods as indicated.

These findings suggest strongly that upregulation of p27^kip1 is involved in IFN--mediated inhibition of bronchial epithelial cell proliferation.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Bronchial epithelial cells are exposed to a variety of noxious agents such as bacteria, viruses, tobacco smoke, and air pollutants. These exogenous agents, as well as endogenous factors including neutrophil elastase and eosinophil cationic protein, can cause mucosal injury, which is followed by a series of repair processes. Plasma-derived fibrin covers the injured sites, to which bronchial epithelial cells migrate, proliferate, and differentiate, and thereby the repair processes are completed. This compensatory proliferation of bronchial epithelial cells is essential for homeostasis of the airway and lung regeneration and is considered to be regulated both positively and negatively by various growth factors and cytokines. Dysregulated growth and differentiation of bronchial epithelial cells will lead to airway remodeling, found in chronic bronchial asthma. Therefore, it is important to elucidate the molecular basis of the regulation of human bronchial epithelial cell proliferation (17).

HGF, which was originally identified as a potent mitogen for rat hepatocyte in primary culture (1), is now considered to be a mesenchyme-derived pleiotropic factor that stimulates cell growth, cell motility, and morphogenesis in a wide range of tissues and organs. HGF exerts its various effects through a high affinity HGF receptor, c-Met. Binding of HGF causes autophosphorylation of tyrosine residues of c-Met, which interact with Src homology 2 domains of several intracellular proteins. The serine/threonine protein kinase cascade consisting of Raf, MAP kinase kinase (MEK), and MAP kinase (ERK) is one of the well-characterized Ras effectors (18). Upon activation, ERK translocate to the nucleus and regulate gene expression by phosphorylating transcription factors such as Elk-1 and AP-1. It has been reported that HGF induces activation of ERK in human melanocytes, mouse keratinocytes, rat hepatocytes, rat alveolar type II cells (19), Madin-Darby canine kidney epithelial cells, human stomach adenocarcinoma cells, and human corneal epithelial cells. Activation of MAP kinase in various types of cells mediates cell proliferation, scattering, and differentiation.

Accumulating evidence indicates that HGF acts as a pulmotrophic factor. Ohmichi and coworkers have demonstrated recently that HGF acted as a morphogenic factor in fetal lung (28). HGF stimulated DNA synthesis in adult rat alveolar type II cells in vitro (6). Intratracheal administration of hydrochloride induced acute lung injury in rats, which was followed by augmented HGF production and increased DNA synthesis in alveolar type II cells (5). Clinical studies have also demonstrated that patients with inflammatory lung diseases show elevated concentrations of HGF in sera and bronchoalveolar lavage fluids (21). Immunohistochemical studies have shown that the source of HGF included alveolar macrophages and interstitial fibroblasts (22). However, it remains unknown whether HGF has any direct effect on human bronchial epithelial cells.

We studied the effect of HGF on proliferation of human peripheral bronchial epithelial cells that were obtained from healthy adult volunteers with an ultrathin fiberoptic bronchoscope (10, 11). We also studied the molecular basis for the mitogenic action of HGF by using BEAS-2B cells, an immortalized cell line derived from human bronchial epithelial cells. We found that HGF stimulated proliferation of normal human bronchial epithelial cells and BEAS-2B cells (Figures 1A and 1C). The mitogenic effect of HGF was dose-dependent, with a maximally effective concentration being approximately 50 ng/ml (Figures 1B and 1D). Importantly, the effect of HGF on cell proliferation depended absolutely on costimulation with serum; in the absence of serum, HGF did not stimulate proliferation of either human normal bronchial epithelial cells or BEAS-2B cells. However, HGF by itself induced a rapid and sustained activation of ERK in normal human bronchial epithelial cells and BEAS-2B cells (Figure 2). The MEK1 inhibitor PD98059 inhibited HGF-induced activation of ERK (Figure 4A) and proliferation completely in response to a combination of HGF and serum (Figure 4B). These composite findings demonstrated that the MEK- ERK-dependent pathway is necessary but not sufficient for mitogenesis of human bronchial epithelial cells. It has been reported that the growth-stimulating potential of HGF is dependent on the coexistence of serum in rat alveolar type II epithelial cells (6). We also studied the effect of LY294002, a specific inhibitor of PI 3-kinase, on HGF-induced cell proliferation of BEAS-2B cells. LY294002 inhibited HGF-induced cell proliferation partially, without inhibiting HGF-induced ERK activation. These findings indicate that the ERK- and the PI 3-kinase signaling pathways are independently involved in HGF-induced proliferation of human bronchial epithelial cells. We found that PD98059 (1-100 µM) did not show any inhibition of HGF-stimulated activation of Akt, or anisomycin-stimulated activation of p38 MAP kinase or JNK, indicating that PD98059 selectively inhibited the MEK-ERK pathway, and that it is not cytotoxic. The results obtained from pharmacologic studies should be confirmed by other experimental approaches including employment of dominant negative forms of proteins or cells, with targeted disruption of specific genes, to elucidate the roles of these pathways further.

Cell cycle progression is regulated by a family of serine/ threonine protein kinases termed CDK, which are both positively and negatively regulated by association with coactivators, cyclins, and CDK inhibitors, respectively (23, 24). Progression through the G1 phase of the mammalian cell cycle is mediated through an early induction of D-type cyclins, including cyclin D1 (25), and downregulation of p27^kip1 CDK inhibitor. In this study, we demonstrated that HGF by itself induced the protein expression of cyclin D1 in the human bronchial epithelial cell line BEAS-2B. Recently, the relationship between the Ras/ERK signaling pathway and cell cycle progression has been studied extensively. Several reports showed that the Ras/ERK signaling pathway is involved in the induction of cyclin D1 and in the downregulation of p27^kip1 using the dominant-negative Ras mutant (26). Platelet-derived growth factor and epidermal growth factor induce cyclin D1 expression via the Ras/ERK signaling pathway in airway smooth muscle cells and fibroblasts (27). We examined the role of ERK in HGF-induced cyclin D1 expression using PD98059 and found that PD98059 did not inhibit HGF-induced cyclin D1 expression in human bronchial epithelial cells. It is probable that PD98059 might inhibit HGF-induced phosphorylation of ERK only transiently or that signals other than ERK might be important in HGF-induced expression of cyclin D1. Indeed, it was reported that PI 3-kinase mediated induction of cyclin D1 and that p38 MAP kinase downregulated the expression of cyclin D1 (28, 29).

IFN- is a negative growth regulatory cytokine for Th2 lymphocytes and some nonlymphoid cells, including epithelial cells. We have demonstrated recently that IFN- inhibits proliferation of human bronchial epithelial cells (7). In the present study, we confirmed our previous results, demonstrating that IFN- inhibited HGF-stimulated proliferation of human bronchial epithelial cells obtained from healthy volunteers in a dose-dependent manner (Figure 5). Further, we studied whether or not IFN- interfered with HGF-induced mitogenic signaling. IFN- did not inhibit HGF-induced activation of ERK (Figure 6A) or the expression of cyclin D1 (Figure 6B) in human bronchial epithelial cells. We speculated that IFN- inhibited HGF-induced cell proliferation by upregulating the cellular levels of one or more CDK inhibitors. Indeed, it has been reported that IFN- induces the expression of p21^WAF1/CIP1 in hepatocytes, and p27^kip1 in mammary epithelial cells (30). We found that IFN- induced the expression of p27^kip1 in human bronchial epithelial cells in the presence of HGF and serum (Figure 7C). These findings suggest strongly that p27^kip1, at least in part, is involved in the IFN- inhibition of HGF-induced cell proliferation. It is also known that TGF-, another potent growth inhibitor, induces the expression of p27^kip1 and p21^WAF1/CIP1 CDK inhibitors in various cell types (31). Previous studies have shown that either the rate of synthesis of p27^kip1 or the rate of degradation of the protein, or both, are involved in the regulation of the p27^kip1 levels by the extracellular growth signals (26, 32). Definitely, the mechanism by which IFN- upregulates the p27^kip1 protein level in human bronchial epithelial cells is the subject of our future investigation.

In conclusion, HGF stimulated proliferation of human bronchial epithelial cells in the presence of serum through mechanisms involving ERK1/2 and PI 3-kinase. In the absence of serum, HGF was capable of activating ERK and upregulating the expression level of cyclin D1, which, however, were not sufficient for epithelial cell proliferation. IFN- inhibited HGF- and serum-stimulated proliferation of human bronchial epithelial cells, without inhibiting the effect of HGF on ERK or cyclin D1 induction. Rather, IFN- upregulated the expression level of p27^kip1 CDK inhibitor in the presence of HGF and serum potently. Further studies are required to elucidate the roles of HGF and IFN- in the regeneration and remodeling of airway epithelial cells after bronchial mucosal injury.