Defect of Pro-Hepatocyte Growth Factor Activation by ... - ATS Journals

Defect of Pro-Hepatocyte Growth Factor Activation by Fibroblasts in Idiopathic Pulmonary Fibrosis Sylvain Marchand-Adam, Aure´lie Fabre, Arnaud Andre´ Mailleux, Joelle Marchal, Christophe Quesnel, Hiroaki Kataoka, Michel Aubier, Monique Dehoux, Paul Soler, and Bruno Crestani Inserm Unit 700, Institut National de la Sante´ et de la Recherche Medicale, Faculte´ Xavier Bichat; Universite´ Paris 7, Denis Diderot; Service de ˆ pitaux de Paris; Laboratoire de Biochimie A, Ho ˆ pital Bichat, Paris, France; Department of Cell Biology, Pneumologie, Assistance Publique–Ho Harvard Medical School, Boston, Massachusetts; and Faculty of Medicine, University of Miyazaki, Miyazaki, Japan

Rationale and Objectives: Hepatocyte growth factor (HGF) protects against lung fibrosis in several animal models. Pro-HGF activation to HGF is subjected to regulation by its activator (HGFA), a serine protease, and HGFA-specific inhibitors (HAI-1 and HAI-2). Our hypothesis was that fibroblasts from patients with idiopathic pulmonary fibrosis (IPF) had an altered capacity to activate pro-HGF in vitro compared with control fibroblasts. Methods: We measured the kinetics of pro-HGF activation in human lung fibroblasts from control subjects and from patients with IPF by Western blot. HGFA, HAI-1, and HAI-2 expression was evaluated by immunohistochemistry, RNA protection assay, and Western blot. We evaluated the effect of TGF-␤1 and PGE2 on pro-HGF activation and HGFA, HAI-1, and HAI-2 expression. Main Results: Lung fibroblasts activated pro-HGF in vitro. Pro-HGF activation was inhibited by serine protease inhibitors, by an antiHGFA antibody, as well as by HAI-1 and HAI-2. Pro-HGF activation by IPF fibroblasts was reduced compared with control fibroblasts. In IPF fibroblasts, HGFA expression was lower and HAI-1 and HAI-2 expression was higher compared with control fibroblasts. PGE2 stimulated pro-HGF activation through increased expression of HGFA and decreased expression of its inhibitor HAI-2. In contrast, TGF-␤1 reduced the ability of lung fibroblasts to activate pro-HGF through decreased expression of HGFA and increased expression of its inhibitors. Conclusions: IPF fibroblasts have a low capacity to activate pro-HGF in vitro via a low level of HGFA expression and high levels of HAI-1 and HAI-2 expression, and PGE2 is able to partially correct this defect. Keywords: human; lung; prostaglandin E2; serine protease; usual interstitial pneumonia

Hepatocyte growth factor (HGF) plays a key role in lung homeostasis, particularly in the alveolus (1, 2). HGF limits lung fibrosis in vivo after bleomycin injury in rodents when given intravenously (3) or intratracheally (4), whereas the use of an inhibitory anti-HGF antibody worsens pulmonary fibrosis (5). The antifibrotic effect of HGF is thought to be mediated by its facilitation of epithelial repair that follows lung injury (6). Inhibition of the

(Received in original form July 12, 2005; accepted in final form March 20, 2006 ) Supported by a grant from the Fondation pour la Recherche Me´dicale (Prix Mariane Josso) and the Fondation Benaid (S.M.-A.). Part of this research program benefited by a grant from the Fondation de France and has been supported by a grant from the Chancellerie des Universite´s de Paris (Legs Poix). P.S. is the recipient of a Contrat d’Interface INSERM-Assistance Publique/Ho ˆ pitaux de Paris. Correspondence and requests for reprints should be addressed to Bruno Crestani, M.D., Service de Pneumologie, Ho ˆ pital Bichat, 16 rue Henri Huchard, 75877 Paris, Cedex 18, France. E-mail: [email protected] This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org Am J Respir Crit Care Med Vol 174. pp 58–66, 2006 Originally Published in Press as DOI: 10.1164/rccm.200507-1074OC on March 30, 2006 Internet address: www.atsjournals.org

fibrotic process by HGF has been reported in other organs such as the liver or the kidney (7). HGF is secreted by various cell types, including pulmonary fibroblasts, as an inactive 92-kD single-chain precursor form, proHGF (8, 9), and remains in this form under physiologic conditions. Once tissue injury occurs, active mature HGF is produced from pro-HGF by proteolytic processing. Mature HGF is a heterodimeric molecule consisting of an ␣ chain (62 kD) and a ␤ chain (32–36 kD) held together by a disulfide bond. To be biologically active, its heterodimeric form is required (9–11). Therefore, the regulation of proteolytic activation of pro-HGF appears to be an important event in the regulation of HGF biological activity. Among the serine proteases able to cleave pro-HGF, HGF activator (HGFA) is the most efficient both in vitro (8, 12) and in injured tissues (9). HGFA is synthesized as a 96-kD inactive precursor and is secreted by various cell types such as hepatocytes (8). HGFA expression may be increased in the injured organ (9). HGFA is then locally activated to the active form (34 kD), mainly by thrombin (13). The bioactivity of HGFA is regulated by its specific inhibitors: HGF activator inhibitor type 1 (HAI-1) and HGF activator inhibitor type 2 (HAI-2) (14, 15). HAI-1 and HAI-2 are structurally similar and each contains two Kunitz-type serine proteinase inhibitor domains (14, 15). Both HAI-1 and HAI-2 genes are abundantly expressed in various tissues including the lung (16). HAI-1 protein has been localized on the cellular lateral surface (17), whereas HAI-2 protein has been observed mainly in the cytoplasm or at the apical surface (18). Accordingly, the activation of pro-HGF is regulated in vivo by the dynamic interactions of HGFA and its inhibitors. Despite the critical importance of HGF in lung homeostasis and its well-identified antifibrotic properties, pro-HGF activation and its regulation have never been evaluated in lung fibrosis. Our group has previously shown that fibroblasts from patients with idiopathic pulmonary fibrosis (IPF) secrete low levels of HGF in vitro (19). We hypothesized that low HGF expression by IPF fibroblasts could be associated with an imbalance in the pro-HGF activating process. Therefore, we examined the proHGF activation pathway and the expression of its main regulatory molecules, HGFA, HAI-1, and HAI-2, in pulmonary fibroblasts cultured from patients with IPF and control patients. Some of the results of these studies have been previously reported in the form of an abstract (20).

METHODS This study was approved by the ethics committee of Paris-Bichat University Hospital (Paris, France).

Patients with Lung Fibrosis Fibroblasts were derived from lung tissue sampled from 10 patients with IPF. Lung samples were obtained by open lung biopsy (n ⫽ 5) or at the time of lung transplantation (n ⫽ 5). IPF was diagnosed according to American Thoracic Society/European Respiratory Society consensus

Marchand-Adam, Fabre, Mailleux, et al.: Pro-HGF Activation in IPF

criteria (21). Patients (eight males and two females) had a median age of 50 yr (range, 44–69 yr). Seven were ex-smokers and three had never smoked. At the time of lung sampling, three patients were being treated with low-dose oral corticosteroids, associated with azathioprine in one patient.

Control Patients Fibroblasts were derived from lung samples from six patients (four males and two females) undergoing lung surgery for removal of a primary lung tumor. Normal lung sample was obtained from a noninvolved segment, remote from the solitary lesion. The median age of these patients was 54 yr (range, 28–68 yr). Two patients had never smoked, two were ex-smokers, and two were active smokers.

Culture of Fibroblasts Human lung fibroblasts were cultured from lung explants until passage 5 as previously described (19).

Pro-HGF Activation by Human Fibroblasts Pro-HGF activation was assayed by Western blot, measuring the proHGF and cleaved HGF forms in fibroblast supernatants as described previously (22, 23) (see the online supplement for additional details). In brief, the fibroblasts were incubated in 50 ␮l of medium containing recombinant human HGF (rhHGF) consisting of a mix of pro-HGF and mature HGF (1 ␮g/ml) for various durations (0 min, 15 min, 30 min, 1 h, 3 h, and 5 h) in the presence of thrombin (5 units/ml) for HGFA activation. To explore the modulation of the pro-HGF activation process, the following substances were added 30 min before adding rhHGF: (1 ) a mix of serine protease inhibitors (10⫺5 M aprotinin, 10⫺4 M leupeptin, and 10⫺3 M phenylmethylsulfonyl fluoride), (2 ) recombinant mouse HAI-1 (4 ␮g/ml), (3 ) recombinant human HAI-2 (4 ␮g/ml), (4 ) antiHGFA polyclonal antibody (20 ␮g/ml), (5 ) anti–HAI-1 (100 ␮g/ml) or anti-HAI-2 (100 ␮g/ml), two mouse monoclonal antibodies (14, 15), or (6 ) control IgG monoclonal antibodies (100 ␮g/ml). In some experiments, fibroblasts were incubated for 18 h with recombinant human transforming growth factor-␤1 (TGF-␤1; 10 ng/ml), prostaglandin E2 (PGE2; 10⫺6 M), or indomethacin (2.5 ␮g/ml) before adding rhHGF. In some experiments, we incubated pro-HGF with conditioned medium from control (n ⫽ 4) and IPF fibroblasts (n ⫽ 4) in the presence of thrombin, for 30 min to 5 h. Conditioned medium consisted of Dulbecco’s modified Eagle’s culture medium without fetal bovine serum, recovered after a 1-h incubation with fibroblasts. HGF Western blot analysis was performed under reducing conditions as previously described (24) (see the online supplement for additional details). Results of pro-HGF activation were expressed as the ratio of remaining pro-HGF relative to pro-HGF measured at 0 min.

Immunohistochemical Detection of HGFA, HAI-1, and HAI-2 We detected HGFA, HAI-1, and HAI-2 expression in control and IPF fibroblasts by using a polyclonal goat antibody to HGFA and monoclonal mouse antibodies to HAI-1 and HAI-2 (see the online supplement for additional details).

Quantitative Analysis of HGFA, HAI-1, and HAI-2 mRNA HGFA, HAI-1, and HAI-2 mRNAs were quantified by real-time reverse transcriptase-polymerase chain reaction. The transcripts of human 60S ribosomal protein L13A (RPL13A) served as endogenous RNA control samples (24) (see the online supplement for additional details). Total RNA was extracted from fibroblasts as previously described (24). Results were expressed as the ratio of HGFA, HAI-1, or HAI-2 to RPL13A.

Western Blot Analysis of HGFA and HAI-1 HGFA and HAI-1 protein content was evaluated by Western blot (see the online supplement for additional details). Results for HGFA and HAI-1 were expressed as the ratio of HGFA or HAI-1 to ␤-actin.

59

Statistical Analysis All data were expressed as the median (range). Differences between IPF and control groups were determined by Mann-Whitney U test. To compare the effects of different agents under baseline conditions, we used the Wilcoxon paired nonparametric test for group comparisons. Correlations were assessed by Spearman rank-order test. p ⬍ 0.05 was considered significant.

RESULTS Pro-HGF Activation by Lung Fibroblasts

We have previously observed that HGF was present mainly in the cleaved mature form (presence of the 69-kD ␣ chain) in IPF and control lung fibroblast supernatants after 18 h of culture (19). We studied the ability of IPF and control fibroblasts to activate a fixed pro-HGF quantity added to the cell culture medium. In all control cultures, exogenous pro-HGF was quickly cleaved to mature HGF. After a 1-h incubation, 17% (6 to 50%) of pro-HGF remained (Figures 1A and 1C) and 15% (0 to 33%) remained after 5 h of incubation. Pro-HGF processing was inhibited by 51% (35 to 76%) by a mix of serine protease inhibitors, thus demonstrating the partial dependence of the activation process on serine proteases in fibroblast cultures (Figure 1A). In the absence of exogenous HGF, pro-HGF and HGF were not detected in fibroblast supernatants after 5 h of incubation (Figure 1A). Pro-HGF activation in all IPF fibroblast cultures was much slower than in control fibroblasts. In IPF fibroblasts, remaining pro-HGF was 80% (64 to 84%) and 49% (30 to 55%) after 1 and 5 h of incubation, respectively (p ⬍ 0.05, compared with control subjects; Figures 1B and 1C). Fibroblast-conditioned medium had a small activating effect on exogenous pro-HGF in the presence of thrombin but without any significant difference between IPF and control conditioned media (see Figure E1 of the online supplement for additional details). We assessed whether the HGFA-dependent pathway of proHGF activation was involved in pro-HGF activation in vitro. Inhibition of HGFA consistently stopped pro-HGF activation by control fibroblasts. Indeed, in control fibroblasts, addition of a neutralizing anti-HGFA antibody reduced the activation of pro-HGF by 31% (2 to 53%) (p ⬍ 0.05), and addition of recombinant mouse HAI-1 or recombinant human HAI-2 inhibited this process by 35% (19 to 61%) and 27% (2 to 52%), respectively (p ⬍ 0.05; Figures 2A and 2B). The neutralizing anti–HAI-1 and anti–HAI-2 antibodies did not modify pro-HGF activation. These results demonstrate that the HGFA-dependent pathway of pro-HGF activation is active in lung fibroblasts in vitro. In similar experiments performed with IPF fibroblasts, antiHGFA antibody also partially inhibited pro-HGF activation (90% [42 to 100%] vs. 64% [28 to 85%] of pro-HGF remaining at 1 h) (Figures 2C and 2D). In contrast, addition of neutralizing anti–HAI-1 antibody increased the rate of pro-HGF activation (46% [2 to 66%] vs. 64% [28 to 85%] of pro-HGF remaining at 1 h). Anti–HAI-2 antibody had a similar positive effect (31% [4 to 50%] of remaining pro-HGF). The mouse and goat IgG control antibodies had no effect on pro-HGF activation (see Figure E2 for additional details). Altogether, these results show that lung fibroblasts are able to activate pro-HGF by HGFA-dependent cleavage. They also point toward a slowdown in pro-HGF activation in IPF fibroblast cultures, which is modulated by HGFA activation or inhibition. HGFA Expression in Lung Fibroblasts in vitro

In the next set of experiments, we hypothesized that a difference in the level of expression of HGFA or its physiologic inhibitors,

60

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 174 2006

Figure 1. Ability of idiopathic pulmonary fibrosis (IPF) and control fibroblasts to activate pro-hepatocyte growth factor (pro-HGF). Shown is a representative Western blot of the response over time of pro-HGF activation in control fibroblasts (A ) and in IPF fibroblasts (B ). A mix of proHGF and mature HGF was added to cell cultures for 0 to 5 h. At the end of the incubation period, Western blot analysis was performed. Pro-HGF (92 kD) was quickly cleaved to the mature heterodimer HGF (␣-chain HGF, 62 kD and ␤-chain HGF, 32–36 kD). Pro-HGF processing was inhibited by a mix of serine protease inhibitors (SPI mixed). We never detected HGF after 5 h of incubation in the supernatant of IPF and control fibroblast cultures without exogenous pro-HGF. (C ) In IPF fibroblasts (n ⫽ 4, solid circles), pro-HGF activation was much slower that in control fibroblasts (n ⫽ 4, open circles). Results are expressed as the percentage of remaining pro-HGF as a function of time.

HAI-1 and HAI-2, could explain the slow rate of pro-HGF activation by IPF fibroblasts. We first evaluated HGFA expression in fibroblasts. In vitro, every fibroblast cultured from both normal and fibrotic lung strongly expressed HGFA, as assessed by immunohistochemistry (Figures 3A and 3B). HGFA mRNA was detected in all lung fibroblasts. The HGFA:RPL13A mRNA ratio was lower in IPF fibroblasts (0.15 [0.13 to 0.21]) compared with control fibroblasts (0.30 [0.12 to 0.38]; p ⬍ 0.05; Figure 3C). Moreover, in the cell lysates, the content of the active form of HGFA was lower in IPF fibroblasts (0.56 [0.32 to 0.73]) compared with control fibroblasts (1.00 [0.92 to 1.16]; p ⬍ 0.05; Figures 3D and 3E). Interestingly, the active form of HGFA was not detected in IPF and control fibroblast supernatants (data not shown).

Figure 2. Role of endogenous hepatocyte growth factor activator (HGFA) in pro-HGF activation in IPF and control fibroblasts. (A ) Representative Western blot of pro-HGF in control fibroblasts after 1 h of incubation with anti-HGFA, anti-HGF activator inhibitor type 1 (anti–HAI-1), anti-HGF activator inhibitor type 2 (anti–HAI-2), recombinant mouse HAI-1 (rmHAI-1), or recombinant human HAI-2 (rhHAI-2). (B ) Remaining pro-HGF after 1 h of incubation (expressed as a percentage of pro-HGF at time 0) in control fibroblasts (n ⫽ 5). Anti-HGFA, rmHAI-1, and rhHAI-2 inhibited pro-HGF activation. (C ) Representative Western blot of pro-HGF in IPF fibroblasts (n ⫽ 5) after 1 h of incubation with anti-HGFA, anti–HAI-1, anti–HAI-2 monoclonal antibodies, rmHAI-1, or rhHAI-2. (D ) Percentage of pro-HGF remaining after a 1-h incubation with IPF fibroblasts (n ⫽ 5). Anti-HGFA inhibited pro-HGF activation. In contrast, addition of neutralizing anti–HAI-1 or anti–HAI-2 increased the rate of pro-HGF activation. Results of pro-HGF activation are expressed as a ratio of pro-HGF remaining relative to pro-HGF at 0 min. † p ⬍ 0.05 compared with 1 h.

These results demonstrate a defect in the expression of HGFA expression in IPF fibroblasts. HAI-1 and HAI-2 Expression by Lung Fibroblasts in vitro

After its activation, the action of HGFA on pro-HGF activation is negatively regulated by two serine protease inhibitors, HAI-1 and HAI-2 (14, 15). We therefore evaluated HAI-1 and HAI-2 expression in IPF and control fibroblast cultures.


61

Figure 3. HGFA expression in IPF and control fibroblasts. Control fibroblasts (A ) and IPF fibroblasts (B ) were strongly labeled with the antiHGFA antibody. (C ) HGFA mRNA expression in control (n ⫽ 5) and IPF (n ⫽ 8) fibroblasts. The transcripts of human RPL13A served as endogenous RNA controls. Results are expressed as the HGFA:RPL13A ratio. The HGFA:RPL13A mRNA ratio was lower in IPF fibroblasts than in control fibroblasts (p ⬍ 0.05). (D ) Representative Western blot of cell-lysate active HGFA for three different control (C1, C2, and C3) and three different IPF (IPF1, IPF2, and IPF3) cultures. ␤-Actin was used as an internal control. (E ) Active HGFA content, expressed as the HGFA:␤actin ratio. Active HGFA content in cell lysates was lower in IPF fibroblasts than in control fibroblasts (p ⬍ 0.05). Individual values and median (bar) are shown. †p ⬍ 0.05 compared with control fibroblasts.

Figure 4. HAI-1 and HAI-2 expression in IPF and control fibroblasts. HAI-1 and HAI-2 expression was detected by immunohistochemistry in all control fibroblasts (A and C ) and IPF fibroblasts (B and D ). (E ) HAI-1 and HAI-2 mRNA expression in IPF fibroblasts (n ⫽ 10) and control fibroblasts (n ⫽ 6). Transcripts of human RPL13A served as endogenous RNA controls. HAI-1:RPL13A and HAI-2:RPL13A mRNA ratios were higher in IPF fibroblasts than in control fibroblasts (p ⬍ 0.05). (F ) Representative Western blot of HAI-1 from fibroblast samples of two different control subjects (C1 and C2) and two different patients with IPF (IPF1 and IPF2). The hepatocyte cell line Hep3b, known to produce HAI-1, served as a positive control (41). ␤-Actin was used as an internal control. (G ) HAI-1 content, expressed as HAI-1:␤-actin ratio. HAI-1 content in cell lysates was higher in IPF fibroblasts (n ⫽ 8) than in control fibroblasts (n ⫽ 5; p ⬍ 0.05). Individual values and median (bar) are shown. †p ⬍ 0.05 compared with control fibroblasts.

HAI-1 and HAI-2 expression was detected by immunohistochemistry in every fibroblast, both in control subjects and in IPF (Figures 4A–4D). HAI-1 and HAI-2 mRNAs were detected in all lung fibroblasts. The HAI-1:RPL13A and HAI-2:RPL13A mRNA ratios were higher in IPF fibroblasts (0.05 [0.01 to 0.50] and 0.10 [0.03 to 2.25], respectively) compared with control fibroblasts (0.01 [0.004 to 0.02], and 0.02 [0.01 to 0.79], respectively; p ⬍ 0.05 for both comparisons; Figure 4E).

In Western blot studies, HAI-1 expression was higher in IPF fibroblasts (0.19 [0.12 to 0.28]) than in control fibroblasts (0.15 [0.08 to 0.19]; p ⬍ 0.05; Figures 4F and 4G). HAI-1 was not detected in IPF and control fibroblast supernatants (data not shown). These results demonstrate an increase in the expression of HAI-1 and HAI-2, two inhibitors of HGFA, and a decrease in HGFA expression by IPF fibroblasts, all contributing to a slower pro-HGF activation process.

62


Figure 5. Effect of prostaglandin E2 (PGE2) stimulation on HGFA, HAI-1, and HAI-2 mRNA expression in control (A ) and IPF (B ) fibroblasts. Fibroblasts were cultured for 18 h in the presence of PGE2 (10⫺6 M) and HGFA, HAI-1, and HAI-2 mRNA expression was quantified by reverse transcription–polymerase chain reaction (RT-PCR). PGE2 increased HGFA and HAI-1 mRNAs and decreased HAI-1 and HAI-2 mRNAs in control and IPF fibroblasts, respectively (p ⬍ 0.05). Results are expressed as a percentage of baseline HGFA, HAI-1, or HAI-2 mRNA expression. Individual values and median (bar) are shown. †p ⬍ 0.05 compared with unstimulated conditions.

PGE2 Stimulates Pro-HGF Activation

We and others have previously observed that secretion of PGE2, an antifibrotic mediator (25), by IPF fibroblasts is profoundly decreased and could contribute to reduced HGF secretion (19). We evaluated whether PGE2 could favorably regulate pro-HGF activation by lung fibroblasts through the modulation of HGFA, HAI-1, and HAI-2 expression. PGE2 increased HGFA mRNA by 80% (22 to 123%; p ⬍ 0.05), but PGE2 had contrasting effects on the expression of its inhibitors. PGE2 increased HAI-1 mRNA by 170% (154 to 330%) but decreased HAI-2 mRNA by 45% (⫹80 to –88%) in control fibroblasts (p ⬍ 0.05; Figure 5A). PGE2 had a similar effect on HGFA (⫹42% [–29 to ⫹120%]), HAI-1 (⫹75% [–36 to ⫹259%]), and HAI-2 (–50% [–80 to ⫹73%]) expression in IPF fibroblasts (p ⬍ 0.05 for all comparisons; Figure 5B). We next evaluated the kinetics of pro-HGF activation in the presence of PGE2. Culture of control fibroblasts with PGE2 (10⫺6 M) for 18 h had no significant effect on the pro-HGF activation capacity of fibroblasts, most probably because pro-HGF activation was already maximal (Figures 6A and 6B). In contrast, incubation of IPF fibroblasts with PGE2 for 18 h increased proHGF activation capacity by 36% (17 to 63%; p ⬍ 0.05; Figures 6C and 6D). The effect of PGE2 was completely inhibited by incubation with anti-HGFA antibody, indicating that pro-HGF activation modulation by PGE2 involves HGFA. The ranges of response to PGE2 were similar in IPF and control fibroblasts.

Figure 6. Effect of PGE2 stimulation on pro-HGF activation in IPF and control fibroblasts. Fibroblasts were cultured for 18 h in the presence of PGE2. A mix of pro-HGF and mature HGF was then added to the cell culture for 1 h. Pro-HGF was analyzed by Western blot. (A and C ) Representative Western blots of remaining pro-HGF in control (A ) and IPF (C ) fibroblast cultures with PGE2 and/or anti-HGFA antibody. PGE2 had no effect on the capacity of fibroblasts to activate pro-HGF (B ). In contrast, PGE2 increased the pro-HGF activation capacity of IPF fibroblasts (D ). The effect of PGE2 was completely inhibited by incubation with an anti-HGFA antibody (D ). Remaining pro-HGF was expressed relative to pro-HGF at time 0. †p ⬍ 0.05 compared with 1 h.

To evaluate the effect of endogenous prostaglandins on the pro-HGF activation process, control fibroblasts were incubated for 18 h with indomethacin, an inhibitor of cyclooxygenase-1 and cyclooxygenase-2. Indomethacin reduced pro-HGF activation by 53% (24 to 89%; p ⬍ 0.05; Figure 7A) and strongly decreased HGFA mRNA expression by 88% (38 to 99%; p ⬍ 0.05; Figure 7B). In contrast, indomethacin had no effect on IPF fibroblasts (Figures 7C and 7D). These results suggest that PGE2 may positively regulate the pro-HGF activation process by IPF fibroblasts essentially through an increase in HGF activator. TGF-␤1 Inhibits Pro-HGF Activation

TGF-␤1 is a major player in both initiation and progression of pulmonary fibrosis. TGF-␤1 has been shown to regulate transcriptional and posttranscriptional expression of HGF in lung


63

Figure 8. Effect of transforming growth factor-␤1 (TGF-␤1) stimulation on HGFA, HAI-1, and HAI-2 mRNA expression in control (A ) and IPF (B ) fibroblasts. Fibroblasts were cultured for 18 h in the presence of TGF-␤1 (10 ng/ml) and HGFA, HAI-1, and HAI-2 mRNA expression was quantified by RT-PCR. TGF-␤1 decreased HGFA mRNA and increased HAI-1 and HAI-2 mRNAs in control and IPF fibroblasts (p ⬍ 0.05). Results are expressed as a percentage of baseline HGFA, HAI-1, or HAI-2 mRNA expression. Individual values and median (bar) are shown. †p ⬍ 0.05 compared with unstimulated conditions.

Figure 7. Effect of indomethacin on pro-HGF activation. Fibroblasts were cultured for 18 h in the presence of indomethacin. Pro-HGF was then added in the cell culture for 1 h. Pro-HGF was analyzed by Western blot. Results for remaining pro-HGF are expressed as a percentage of pro-HGF at time 0. Indomethacin decreased pro-HGF activation in all control cultures (A ) but not in IPF cultures (C ). †p ⬍ 0.05 compared with 1 h. Indomethacin decreased HGFA mRNA in control fibroblasts (B ) (p ⬍ 0.05) but not in IPF fibroblasts (D ). Indomethacin had no significant effect on HAI-1 and HAI-2 mRNA expression in control fibroblasts (B ) and IPF fibroblasts (D ). Results are expressed as a percentage of baseline HGFA, HAI-1, or HAI-2 mRNA expression. Individual values and median (bar) are shown. †p ⬍ 0.05 compared with unstimulated conditions.

fibroblasts (26, 27). We assessed the ability of TGF-␤1 to modulate pro-HGF activation by influencing HGFA, HAI-1, and HAI-2 expression in lung fibroblasts, and therefore to contribute to the phenotype of IPF fibroblasts. In control fibroblasts, TGF-␤1 decreased HGFA mRNA by 60% (49 to 90%), increased HAI-1 mRNA by 65% (10 to 635%), and increased HAI-2 mRNA by 112% (12 to 428%; p ⬍ 0.05; Figure 8A). TGF-␤1 had a similar effect on HGFA, HAI-1, and HAI-2 mRNA expression in IPF fibroblasts (p ⬍ 0.05; Figure 8B). Incubation with TGF-␤1 decreased pro-HGF activation in all control cultures by 39% (20 to 70%) after 1 h compared with basal conditions (p ⬍ 0.05; Figures 9A and 9B). TGF-␤1 had a

similar effect on IPF fibroblast cultures (74% [37 to 100%] vs. 62% [27 to 85%] of the remaining pro-HGF; Figures 9C and 9D). The range of response to TGF-␤1 was similar in both IPF and control fibroblasts. These results indicate that TGF-␤1 may contribute to the defect of pro-HGF activation by IPF fibroblasts through the modulation of expression of HGF activator, and its specific inhibitors HAI-1 and HAI-2.

DISCUSSION In the present study, we demonstrate for the first time that (1) lung fibroblasts have the ability to activate pro-HGF in vitro through HGFA; (2) IPF fibroblasts have a reduced ability to activate pro-HGF that is related to the decreased expression of HGFA and the increased expression of its inhibitors HAI-1 and HAI-2; (3) PGE2 has the capacity to restore, at least in part, the pro-HGF activation process in IPF fibroblasts; and (4) TGF-␤1 could play a critical role in the inhibition of pro-HGF activation by fibroblasts in IPF. HGF is secreted as an inactive form, pro-HGF, and is associated with the extracellular matrix in the producing tissues. To generate biologically active HGF, the proteolytic conversion of the single-chain precursor form, pro-HGF, to the two-chain heterodimeric active form is essential (10). This proteolytic process is a critical event to regulate HGF activity in the extracellular microenvironment.

64


Figure 9. Effect of TGF-␤1 stimulation on pro-HGF activation process in IPF (n ⫽ 5) and control (n ⫽ 5) fibroblasts. Fibroblasts were cultured for 18 h in the presence of TGF-␤1. A mix of pro-HGF and mature HGF was then added to the cell culture for 1 h. Pro-HGF was analyzed by Western blot. (A and C ) Representative Western blots of remaining proHGF in control (A ) and IPF (C ) fibroblasts (after a 1-h incubation). TGF␤1 decreased pro-HGF activation in all control (B ) and IPF (D ) cultures. Results of remaining pro-HGF are expressed as a percentage of proHGF at time 0. †p ⬍ 0.05 compared with 1 h.

In vitro, HGFA is the most efficient serine protease to cleave pro-HGF to mature HGF (8, 12). Although several other serine proteases are known to activate pro-HGF in vitro, such as urokinase-type activator (28), tissue-type plasminogen activator (28), matriptase I (29), and blood coagulation factors XIIa (12) and XIa (30), as well as plasma kallikrein (30), their role in the activation of pro-HGF in fibroblast cultures is unknown. HGFA activity was first detected in a primary culture of hepatocytes (10). HGFA is expressed mainly in parenchymal cells of adult liver and secreted into the blood. A study on human samples showed that HGFA is also produced in the gastrointestinal tract, is weakly expressed in the kidney, and is almost undetectable in the brain, heart, testis, and ovary (31). To our knowledge, there are no current data on HGFA expression in the lung. Although it has been previously shown that HGFA mRNA was present in a human fetal fibroblast cell line (MRC5) (32), our results demonstrate that HGFA is expressed in human adult lung fibroblasts.

HGFA is secreted as an inactive precursor, and thrombin has been identified as the main and most effective protease for cleavage of the HGFA precursor (13). The activation of HGFA may be neutralized by its two specific serine protease inhibitors, HAI-1 and HAI-2 (14, 15). HAI-1 and HAI-2 bind to HGFA to prevent its binding to pro-HGF and to cleave it into the active heterodimer HGF. The inhibiting abilities of HAI-1/2 rely on the presence of Kunitz domain 1. Both HAI-1 and HAI-2 genes are expressed abundantly in various tissues including the lung (16). To our knowledge, this is the first study to demonstrate pro-HGF activation by human lung fibroblasts. We observed that pro-HGF activation in lung fibroblast cultures was regulated in part by HGFA. We found a decrease in HGFA expression associated with an increase in HAI expression in IPF fibroblasts compared with control fibroblasts. This opposition in HGFA and its inhibitors expression has been previously observed in many cell lines (32). In our study, a mix of serine protease inhibitors, an antiHGFA antibody, or recombinant HAI-1/2 only partially inhibited the pro-HGF activation process, highlighting the partial dependence of the activation process on serine proteases in fibroblast cultures. We cannot explain why this activation is independent of HGFA, but other proteases could activate proHGF in vitro (30). Because we used recombinant mouse HAI-1, interspecies differences may complicate the interpretation of our inhibition assay, as different results might have been obtained with recombinant human HAI-1. However, recombinant human HAI-1 was not available to us. Interestingly, the conditioned medium of fibroblasts had a small capacity to activate pro-HGF compared with fibroblasts. This suggests that pro-HGF activation is essentially controlled by membrane-associated proteases as already shown by Kataoka and colleagues (33). Few studies have evaluated the pro-HGF activation process in the pathophysiology of tissue fibrosis. In a murine model of bleomycin-induced lung fibrosis, pro-HGF was found in lung homogenates but was not found in the bronchoalveolar lavage fluid, indicating that all of the HGF present in bronchoalveolar lavage is activated in this model (5). In a rat cirrhosis model, insufficient HGF activation results in reduced expression of hepatic HGFA and increased expression of splenic HAI-1, and may be the reason for impaired liver regeneration observed after partial hepatectomy compared with normal liver (34). In this model, therapy with exogenous active HGF results in effective liver regeneration via up-regulation of HGFA expression (35). The regulation of HGFA and HAI expression is poorly understood. HGFA and HAI-1 could be upregulated by interleukin 1␤ (33). Studies of HGFA and HAI promoters suggest that cytokines involved in the early inflammatory response could regulate HGFA and HAI-1 expression (36, 37). In the present study, HGFA and HAI expression is modulated by TGF-␤1 and PGE2. TGF-␤1 down-regulates HGFA and upregulates HAI-2 and HAI-1, limiting the pro-HGF activation process. PGE2 has effects opposite to that of TGF-␤1 on HGFA and HAI-2 expression and favors pro-HGF activation. However, both TGF-␤1 and PGE2 stimulate HAI-1 expression. This demonstrates a difference in the regulation of HAI-1 and HAI-2 expression as suggested in previous studies (33, 37). The preferential inhibitory effect of PGE2 on HAI-2 expression is interesting because HAI-2 is a more potent inhibitor of HGFA than is HAI-1 (15). Altogether, our data support the view that part of the protective effect of PGE2 and of the harmful effect of TGF-␤1 on pulmonary fibrosis occurs through the modulation of pro-HGF activation by pulmonary fibroblasts. HGF acts as an antifibrotic molecule, as observed in vivo, where the provision of HGF limits lung fibrosis after bleomycin injury in rodents when given


intravenously (3) or intratracheally (4), even several days after the insult. HGF promotes repair of the alveolar epithelium, acting as a mitogen, motogen (2), and morphogen for alveolar epithelial cells. HGF is also able to protect cells from apoptosis in vitro (38). Reduction of pro-HGF activation by fibroblasts from fibrotic lung may contribute to their ability to induce the apoptosis of alveolar epithelial cells in vitro whereas normal fibroblasts do not (39). This may be of importance as apoptotic alveolar epithelial cells are adjacent to the site of fibroproliferation in vivo in the fibrotic lung (40). We conclude that the capacity for pro-HGF activation of pulmonary fibroblasts obtained from patients with pulmonary fibrosis is profoundly decreased. This defect appears to be secondary to the dysregulation of HGFA and HAI expression and could contribute to the alteration of the alveolar repair process that is a characteristic of the disease. Our study also argues for the role of an imbalance of TGF-␤1 and PGE2 in the pathophysiology of pulmonary fibrosis. Conflict of Interest Statement : None of the authors has a financial relationship with a commercial entity that has an interest in the subject of the manuscript.

References 1. Ware LB, Matthay MA. Keratinocyte and hepatocyte growth factors in the lung: roles in lung development, inflammation, and repair. Am J Physiol Lung Cell Mol Physiol 2002;282:L924–L940. 2. Panos RJ, Patel R, Bak PM. Intratracheal administration of hepatocyte growth factor/scatter factor stimulates rat alveolar type II cell proliferation in vivo. Am J Respir Cell Mol Biol 1996;15:574–581. 3. Yaekashiwa M, Nakayama S, Ohnuma K, Sakai T, Abe T, Satoh K, Matsumoto K, Nakamura T, Takahashi T, Nukiwa T. Simultaneous or delayed administration of hepatocyte growth factor equally represses the fibrotic changes in murine lung injury induced by bleomycin: a morphologic study. Am J Respir Crit Care Med 1997;156:1937– 1944. 4. Dohi M, Hasegawa T, Yamamoto K, Marshall BC. Hepatocyte growth factor attenuates collagen accumulation in a murine model of pulmonary fibrosis. Am J Respir Crit Care Med 2000;162:2302–2307. 5. Hattori N, Mizuno S, Yoshida Y, Chin K, Mishima M, Sisson TH, Simon RH, Nakamura T, Miyake M. The plasminogen activation system reduces fibrosis in the lung by a hepatocyte growth factor–dependent mechanism. Am J Pathol 2004;164:1091–1098. 6. Yamada T, Hisanaga M, Nakajima Y, Mizuno S, Matsumoto K, Nakamura T, Nakano H. Enhanced expression of hepatocyte growth factor by pulmonary ischemia–reperfusion injury in the rat. Am J Respir Crit Care Med 2000;162:707–715. 7. Ueki T, Kaneda Y, Tsutsui H, Nakanishi K, Sawa Y, Morishita R, Matsumoto K, Nakamura T, Takahashi H, Okamoto E, et al. Hepatocyte growth factor gene therapy of liver cirrhosis in rats. Nat Med 1999;5:226–230. 8. Miyazawa K, Shimomura T, Kitamura A, Kondo J, Morimoto Y, Kitamura N. Molecular cloning and sequence analysis of the cDNA for a human serine protease responsible for activation of hepatocyte growth factor: structural similarity of the protease precursor to blood coagulation factor XII. J Biol Chem 1993;268:10024–10028. 9. Miyazawa K, Shimomura T, Kitamura N. Activation of hepatocyte growth factor in the injured tissues is mediated by hepatocyte growth factor activator. J Biol Chem 1996;271:3615–3618. 10. Naka D, Ishii T, Yoshiyama Y, Miyazawa K, Hara H, Hishida T, Kidamura N. Activation of hepatocyte growth factor by proteolytic conversion of a single chain form to a heterodimer. J Biol Chem 1992;267:20114–20119. 11. Weidner KM, Sachs M, Birchmeier W. The Met receptor tyrosine kinase transduces motility, proliferation, and morphogenic signals of scatter factor/hepatocyte growth factor in epithelial cells. J Cell Biol 1993;121: 145–154. 12. Shimomura T, Miyazawa K, Komiyama Y, Hiraoka H, Naka D, Morimoto Y, Kitamura N. Activation of hepatocyte growth factor by two homologous proteases, blood-coagulation factor XIIa and hepatocyte growth factor activator. Eur J Biochem 1995;229:257–261. 13. Shimomura T, Kondo J, Ochiai M, Naka D, Miyazawa K, Morimoto Y, Kitamura N. Activation of the zymogen of hepatocyte growth factor activator by thrombin. J Biol Chem 1993;268:22927–22932.

65 14. Shimomura T, Denda K, Kitamura A, Kawaguchi T, Kito M, Kondo J, Kagaya S, Qin L, Takata H, Miyazawa K, et al. Hepatocyte growth factor activator inhibitor, a novel Kunitz-type serine protease inhibitor. J Biol Chem 1997;272:6370–6376. 15. Kawaguchi T, Qin L, Shimomura T, Kondo J, Matsumoto K, Denda K, Kitamura N. Purification and cloning of hepatocyte growth factor activator inhibitor type 2, a Kunitz-type serine protease inhibitor. J Biol Chem 1997;272:27558–27564. 16. Marlor CW, Delaria KA, Davis G, Muller DK, Greve JM, Tamburini PP. Identification and cloning of human placental bikunin, a novel serine protease inhibitor containing two Kunitz domains. J Biol Chem 1997;272:12202–12208. 17. Kataoka H, Suganuma T, Shimomura T, Itoh H, Kitamura N, Nabeshima K, Koono M. Distribution of hepatocyte growth factor activator inhibitor type 1 (HAI-1) in human tissues: cellular surface localization of HAI-1 in simple columnar epithelium and its modulated expression in injured and regenerative tissues. J Histochem Cytochem 1999;47: 673–682. 18. Kataoka H, Itoh H, Uchino H, Hamasuna R, Kitamura N, Nabeshima K, Koono M. Conserved expression of hepatocyte growth factor activator inhibitor type-2/placental bikunin in human colorectal carcinomas. Cancer Lett 2000;148:127–134. 19. Marchand-Adam S, Marchal J, Cohen M, Soler P, Gerard B, Castier Y, Lese`che G, Valeyre D, Mal H, Aubier M, et al. Defect of hepatocyte growth factor secretion by fibroblasts in idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 2003;168:1156–1161. 20. Marchand-Adam S, Favre A, Marchal J, Quesnel C, Soler P, Kataoka H, Dehoux M, Aubier M, Crestani B. Defect of pro-HGF activation in IPF fibroblast culture [abstract 724]. Eur Respir J 2005;26(Suppl):101. 21. American Thoracic Society. Idiopathic pulmonary fibrosis: diagnosis and treatment. International consensus statement. Am J Respir Crit Care Med 2000;161:646–664. 22. Kataoka H, Hamasuna R, Itoh H, Kitamura N, Koono M. Activation of hepatocyte growth factor/scatter factor in colorectal carcinoma. Cancer Res 2000;60:6148–6159. 23. Tjin EP, Derksen PW, Kataoka H, Spaargaren M, Pals ST. Multiple myeloma cells catalyze hepatocyte growth factor (HGF) activation by secreting the serine protease HGF-activator. Blood 2004;104:2172– 2175. 24. Marchand-Adam S, Plantier L, Bernuau D, Legrand A, Cohen M, Marchal J, Soler P, Leseche G, Mal H, Aubier M, et al. Keratinocyte growth factor expression by fibroblasts in pulmonary fibrosis: poor response to interleukin-1␤. Am J Respir Cell Mol Biol 2005;32:470–477. 25. Keerthisingam CB, Jenkins RG, Harrison NK, Hernandez-Rodriguez NA, Booth H, Laurent GJ, Hart SL, Foster ML, McAnulty RJ. Cyclooxygenase-2 deficiency results in a loss of the anti-proliferative response to transforming growth factor-␤ in human fibrotic lung fibroblasts and promotes bleomycin-induced pulmonary fibrosis in mice. Am J Pathol 2001;158:1411–1422. 26. Matsumoto K, Tajima H, Okazaki H, Nakamura T. Negative regulation of hepatocyte growth factor gene expression in human lung fibroblasts and leukemic cells by transforming growth factor-␤1 and glucocorticoids. J Biol Chem 1992;267:24917–24920. 27. Harrison P, Bradley L, Bomford A. Mechanism of regulation of HGF/ SF gene expression in fibroblasts by TGF-␤1. Biochem Biophys Res Commun 2000;271:203–211. 28. Mars WM, Zarnegar R, Michalopoulos GK. Activation of hepatocyte growth factor by the plasminogen activators uPA and tPA. Am J Pathol 1993;143:949–958. 29. Lee SL, Dickson RB, Lin CY. Activation of hepatocyte growth factor and urokinase/plasminogen activator by matriptase, an epithelial membrane serine protease. J Biol Chem 2000;275:36720–36725. 30. Peek M, Moran P, Mendoza N, Wickramasinghe D, Kirchhofer D. Unusual proteolytic activation of pro-hepatocyte growth factor by plasma kallikrein and coagulation factor XIa. J Biol Chem 2002;277:47804– 47809. 31. Itoh H, Hamasuna R, Kataoka H, Yamauchi M, Miyazawa K, Kitamura N, Koono M. Mouse hepatocyte growth factor activator gene: its expression not only in the liver but also in the gastrointestinal tract. Biochim Biophys Acta 2000;1491:295–302. 32. Parr C, Jiang WG. Expression of hepatocyte growth factor/scatter factor, its activator, inhibitors and the c-Met receptor in human cancer cells. Int J Oncol 2001;19:857–863. 33. Kataoka H, Shimomura T, Kawaguchi T, Hamasuna R, Itoh H, Kitamura N, Miyazawa K, Koono M. Hepatocyte growth factor activator inhibitor type 1 is a specific cell surface binding protein of hepatocyte

66

34.

35.

36.

37.

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 174 2006 growth factor activator (HGFA) and regulates HGFA activity in the pericellular microenvironment. J Biol Chem 2000;275:40453–40462. Kaido T, Oe H, Yoshikawa A, Okajima A, Imamura M. Expressions of molecules associated with hepatocyte growth factor activation after hepatectomy in liver cirrhosis. Hepatogastroenterology 2004;51:547– 551. Xue F, Takahara T, Yata Y, Kuwabara Y, Shinno E, Nonome K, Minemura M, Takahara S, Li X, Yamato E, et al. Hepatocyte growth factor gene therapy accelerates regeneration in cirrhotic mouse livers after hepatectomy. Gut 2003;52:694–700. Miyazawa K, Wang Y, Minoshima S, Shimizu N, Kitamura N. Structural organization and chromosomal localization of the human hepatocyte growth factor activator gene: phylogenetic and functional relationship with blood coagulation factor XII, urokinase, and tissue-type plasminogen activator. Eur J Biochem 1998;258:355–361. Itoh H, Kataoka H, Meng JY, Hamasuna R, Kitamura N, Koono M. Mouse hepatocyte growth factor activator inhibitor type 1 (HAI-1) and

38.

39.

40.

41.

type 2 (HAI-2)/placental bikunin genes and their promoters. Biochem Biophys Acta 2001;1519:92–95. Xiao GH, Jeffers M, Bellacosa A, Mitsuuchi Y, Vande Woude GF, Testa JR. Anti-apoptotic signaling by hepatocyte growth factor/Met via the phosphatidylinositol 3-kinase/Akt and mitogen-activated protein kinase pathways. Proc Natl Acad Sci USA 2001;98:247–252. Uhal BD, Joshi I, True AL, Mundle S, Raza A, Pardo A, Selman M. Fibroblasts isolated after fibrotic lung injury induce apoptosis of alveolar epithelial cells in vitro. Am J Physiol 1995;269:L819–L828. Uhal BD, Joshi I, Hughes WF, Ramos C, Pardo A, Selman M. Alveolar epithelial cell death adjacent to underlying myofibroblasts in advanced fibrotic human lung. Am J Physiol 1998;275:L1192–L1199. Nagata K, Hirono S, Ido A, Kataoka H, Moriuchi A, Shimomura T, Hori T, Hayashi K, Koono M, Kitamura N, et al. Expression of hepatocyte growth factor activator and hepatocyte growth factor activator inhibitor type 1 in human hepatocellular carcinoma. Biochem Biophys Res Commun 2001;289:205–211.