Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Secretory leukoprotease inhibitor (SLPI) is an 11.7-kD, nonglycosylated, single-chain serine protease inhibitor consisting of 107 amino acids (1). It inhibits several serine proteases such as elastase and cathepsin G from neutrophils, trypsin and chymotrypsin from pancreatic acinar cells, and chymase and tryptase from mast cells (1). A major physiologic role of SLPI has been considered as protection against neutrophil elastase (NE) at inflammatory sites. This speculation is based on the biochemical and cell-biologic characteristics of SLPI as follows: (1) a higher association rate constant for NE (~ 10⁷ M¹s¹) than for other proteases (2), (2) its concentration in the local milieu (nearly 10 µM in respiratory epithelial lining fluid) (3, 4), and (3) the cell sources for its production and secretion are restricted to epithelial cells such as secretory cells in respiratory, genital, and lacrimal glands, but do not include the skin, endocrine glands, or hematologic system (5).

However, recent advances in elucidating the new functions of SLPI have revealed that this protein exerts its anti-inflammatory effects through a larger variety of cells than previously thought, and through unexpected functions, some of which are independent of the inhibition of proteases. As for the cell sources, macrophages (6, 7) and neutrophils (6) were shown to express SLPI. Functions independent of the inhibition of proteases include antimicrobial activity (8), inhibition of human immunodeficiency virus-I (HIV-I) infection to lymphocytes (9, 10), and suppression of cyclooxygenase-2 (COX-2) production leading to a reduction of prostaglandin (PG) E₂ and matrix metalloproteinases (MMP-1 and MMP-9) in monocytes (11). Moreover, SLPI is induced by lipopolysaccharide (LPS) or by lipoteichoic acid in macrophages and intracellularly antagonizes the signal transduction pathway elicited by these strong inflammatory stimuli in bacterial cell walls (6, 7, 12).

One of the pharmacologic effects of SLPI against lung inflammation is the suppression of bleomycin (BLM)- induced lung injury (13). One characteristic feature is that SLPI exerts its suppressive effect even at doses that do not inhibit NE from recruited neutrophils (13). This phenomenon also suggests a novel anti-inflammatory activity of SLPI through protease inhibition-independent mechanisms.

To obtain insights about the biologic function of SLPI in the human lung, we tested the hypothesis that human SLPI stimulates the production of hepatocyte growth factor (HGF) in human lung fibroblasts, the major HGF-producing cell type in the lung. HGF was chosen for the study because (1) HGF, otherwise known as scatter factor (SF), is a notable mesenchymal cell-derived cytokine closely implicated in the regulation of mitogenesis, motogenesis, and morphogenesis (14); (2) Rosen and colleagues previously purified a 12-kD HGF-inducing factor from the conditioned media of mouse NIH/3T3 fibroblasts and revealed its N-terminal sequence as AKNDAIKIGA (15), which shows high homology (9 of 10 amino acids) to that of recently cloned mouse SLPI (6, 16, 17); and (3) Yaekashiwa and associates reported the suppressive effect of HGF on BLM-induced lung injury and subsequent pulmonary fibrosis (18).

In the present investigation, we report that human SLPI increases HGF production in human lung fibroblasts by regulating the level of the steady-state messenger RNA (mRNA). We further show that this event depends on the protease- inhibitory activity of SLPI by oxidizing its active center.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Preparation of Recombinant Full- and Half-Length SLPIs

The three recombinant forms of SLPI (full-length SLPI and its N- and C-terminal domains) were produced as described previously (19). Briefly, DNA fragments encoding the amino acids 1-107, 1- 54, and 58-107 of SLPI were synthesized chemically using the appropriate codons from Escherichia coli. The human growth hormone gene was fused to the genes for full- and half-length SLPIs to optimize the expression size via a DNA sequence encoding Leu-Val-Pro-Arg (20), which can be cleaved by thrombin. The expression vectors were constructed and introduced into E. coli HB101. The transformed cells were selected by ampicillin and cultured. The fusion proteins were obtained as inclusion bodies, and were extracted and cleaved with thrombin. The full- and half-length SLPIs were purified by chromatography after S-S refolding. Endotoxin content was determined by limulus amebocyte lysate analysis using KINETIC-QCL (BioWhitaker, Walkersville, MD). The endotoxin content of the full- and half-length SLPIs used in the studies described here were less than 0.6 EU/mg of recombinant protein.

Cell Culture

Human lung fibroblasts (CCD-11Lu, CCD-25Lu, CCD-32Lu, and CCD-33Lu) were purchased from the American Type Culture Collection (Rockville, MD). The human skin fibroblasts used were described previously (21). Cells were grown in Dulbecco's modified Eagle's medium (DMEM) (GIBCO BRL, Gaithersburg, MD) supplemented with 10% fetal calf serum, subcultured at 1:4 using 0.05% trypsin-0.53 mM ethylenediaminetetraacetic acid (GIBCO BRL), and incubated at 37°C in a humidified atmosphere of 5% CO₂ and 95% air. Fibroblasts were studied within five passages after thawing of the ampules.

Measurement of HGF

Human lung and skin fibroblasts were plated on six- or 12-well plates at a density of 5 × 10⁴ cells/cm², and cultured for 24 h. The cells were washed three times with Ca²⁺, Mg²⁺-free phosphate-buffered saline, and re-fed with fresh serum-free DMEM containing various agents: 0-10 µM full- or half-length SLPIs, 50 µg/ml ₁-antitrypsin (₁-AT) (Sigma, St. Louis, MO), 2.5 µg/ml defensin HNP-1 (Sigma), 10 µM ONO-5046 (Ono Pharmaceutical Co., Osaka, Japan), 5 ng/ml interleukin (IL)-1 (Otsuka Pharmaceutical Co., Tokushima, Japan), or 100 ng/ml IL-1 receptor antagonist (IL-1ra) (R&D Systems, Minneapolis, MN). After incubation for a given period of time, the medium was centrifuged to remove debris. The concentration of HGF in the medium was determined by enzyme-linked immunosorbent assay (ELISA) using a Human HGF Immunoassay kit (R&D Systems) and EAR 400 FW spectrophotometer (SLT-LABINSTRUMENTS, Research Triangle Park, NC). The standard curve obtained was linear from 125 to 8,000 pg/ml of HGF.

Northern Blotting

Human lung fibroblast CCD-11Lu cells were plated on 10-cm dishes at a density of 5 × 10⁴ cells/cm². At 24 h after the subcultivation, cells were exposed to 10 µM SLPI for the indicated times or for 12 h at the indicated concentrations of SLPI. Total cellular RNA was then isolated by acid guanidinium thiocyanate-phenol chloroform extraction (22). A total of 20 µg of RNA was electrophoresed on a 1% agarose gel containing 2.2 M formaldehyde and transferred to a nylon membrane. Membranes were probed with the full-length open reading frame complementary DNA (cDNA) for human HGF (21) or the human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA (23) radiolabeled with [-³²P]deoxycytidine triphosphate (~ 111 TBq/mmol) (Du Pont, Wilmington, DE) using the Random Primers DNA Labeling System (GIBCO BRL). After hybridization proceeded for 16 h at 42°C, the membrane was washed, dried, and exposed to a Fuji Imaging Plate (Fuji Photo Film Co., Minamiashigara, Japan). Scanning of the signals was performed using a Bio-Imaging Analyzer system (Fuji Photo Film Co.). The sizes of the transcripts were assessed by comparison to a 0.24-to-9.5-kb RNA Ladder (GIBCO BRL).

Oxidation of SLPI and Its C-Terminal Domain

Full-length SLPI and its C-terminal domain were oxidized with N-chlorosuccinimide (NCS) as described by Boudier and Bieth (24), in which NCS selectively converts surface-exposed methionine residues into methionine sulfoxide derivatives. The reaction mixtures, in a final volume of 62.3 µl, contained 100 mM Tris/HCl, pH 8.3, and a given concentration of NCS with or without 0.64 mM full-length SLPI or 0.64 mM C-terminal domain. After incubation at 25°C for 2 h, the oxidation process was stopped by diluting a 0.55 volume of reaction mixture with a 0.45 volume (51 µl) of 100 mM N-acetylmethionine. The samples were then used directly to assay for HGF-inducing activity.

Statistical Analysis

Statistical analysis was done by the two-tailed Student's t test.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Augmentation of HGF Concentration by SLPI in Culture Media of Human Lung Fibroblasts

When media from four human lung fibroblast cultures (CCD-11Lu, CCD-25Lu, CCD-32Lu, and CCD-33Lu) were assayed for HGF production, all 16-h cultures with 10 µM SLPI showed significant increases in the HGF concentration (Figure 1). Although the sensitivity to SLPI stimuli varied among these fibroblast cell lines, the maximum increase in the HGF concentration was observed in CCD-11Lu cells (5-fold increase compared with SLPI-free control). The three other fibroblast cell lines showed higher HGF concentrations but were less sensitive to the addition of SLPI. No differences in the increases of cell numbers between SLPI-treated and untreated fibroblasts during the 16-h cell culture were found (data not shown).

View larger version (54K): [in this window] [in a new window]   Figure 1.   Effect of SLPI on HGF production by human lung fibroblasts. Four different lines of human lung fibroblasts: CCD-11Lu (A), CCD-25Lu (B), CCD-32Lu (C), and CCD-33Lu (D) were seeded on six-well plates at a density of 5 × 104 cells/cm2 and cultured for 24 h. After the medium was changed to 2 ml of fresh serum-free medium, the cells were cultured for 16 h without (open columns) or with 10 µM full-length SLPI (shaded columns). The concentrations of HGF in the culture medium were determined by ELISA. Results are the means ± standard error of the mean (SEM) for three different wells. Asterisks denote significant difference (P < 0.01) compared with the respective control cells.

Effect of SLPI on HGF mRNA Expression in Human Lung Fibroblasts

Treatment of human lung fibroblast CCD-11Lu cells with SLPI induced a time- and dose-dependent increase in the HGF mRNA expression (Figure 2). The HGF mRNA expression of CCD-11Lu cells was increased at 4 h incubation in the presence of 10 µM SLPI, and was about 4-fold higher at 12 h (Figure 2A). This effect of SLPI was dose-responsive (Figure 2B). Continuous treatment of CCD-11Lu cells with SLPI for 12 h caused a small but detectable increase in HGF transcripts at doses as low as 0.1 µM and a 2-fold increase at 10 µM, whereas no increase was observed without the addition of SLPI. The mRNA levels for HGF corresponded well to the HGF concentrations in the culture media both in the time-course and dose-response experiments (shown as full-length SLPI in Figure 4). These findings further confirm that SLPI induces HGF production by upregulation of the level of the steady-state mRNA.

View larger version (33K): [in this window] [in a new window]   Figure 2.   Effect of SLPI on HGF mRNA levels of human lung fibroblasts. (A) Time course of HGF mRNA expression. CCD-11Lu cells were incubated for various periods (in hours) with 10 µM full-length SLPI in serum-free DMEM. The total cellular RNA was extracted, size-fractionated (20 µg per lane), and Northern blotted. The blot was hybridized to a radiolabeled HGF cDNA probe. Control hybridization with a GAPDH cDNA probe confirmed the integrity of the RNAs used. (B) Correlation of the supernatant SLPI concentration with the HGF mRNA expression. CCD-11Lu cells were incubated for 12 h with full-length SLPI at the indicated concentrations in serum-free medium. Unless otherwise noted, the experimental conditions were the same as for A.

View larger version (16K): [in this window] [in a new window]   Figure 4.   Effect of full- and half-length SLPIs on HGF production. (A) Time course of HGF production by human lung fibroblasts. CCD-11Lu cells were seeded on six-well plates as described in MATERIALS AND METHODS and cultured for 24 h. After the medium was changed to 2 ml of fresh serum-free medium, the cells were cultured for various periods in the absence (control, x) or presence of either 10 µM full-length SLPI (SLPI, filled circles), 10 µM C-terminal domain of SLPI (C-SLPI, filled squares), or 10 µM N-terminal domain of SLPI (N-SLPI, filled triangles). The concentrations of HGF in the culture medium were determined by ELISA. Results are the means ± SEM for three different wells. (B) Dose-responsive production of HGF by human lung fibroblasts. CCD-11Lu cells were cultured for 16 h with full-length SLPI (SLPI, filled circles), C-terminal domain of SLPI (C-SLPI, filled squares), or N-terminal domain of SLPI (N-SLPI, filled triangles) at the indicated concentrations. Unless otherwise noted, the experimental conditions were the same as for A.

SLPI Stimulates HGF Production via an IL-1-Independent Pathway

Whereas HGF production stimulated by IL-1 in human lung and skin fibroblasts was diminished by IL-1ra treatment, the stimulatory effect of SLPI in human lung fibroblasts was not affected by IL-1ra (Figure 3). In this context, the serum-free media from lung fibroblast CCD-11Lu and skin fibroblasts treated with either 10 µM SLPI or 5 ng/ml IL-1 in the presence or absence of 100 ng/ml IL-1ra were assayed for HGF production by ELISA. Although IL-1 significantly stimulated the HGF production of human lung fibroblasts and skin fibroblasts, this effect was markedly inhibited by the simultaneous addition of a 20-fold molar excess of IL-1ra, which competes for binding to the IL-1 receptor without signal transduction. In contrast, the significant increase in HGF production induced by SLPI in lung fibroblasts was not affected by the addition of IL-1ra. In skin fibroblasts, the HGF concentration was neither augmented by the SLPI nor affected by the addition of IL-1ra. These results demonstrate that the stimulation of HGF production by SLPI is dependent on the type of fibroblasts, and that the effect of SLPI is not mediated through IL-1 receptors.

View larger version (16K): [in this window] [in a new window]   Figure 3.   Effect of IL-1ra on the HGF-inducing activity of SLPI in human lung and skin fibroblasts. Human lung fibroblast CCD-11Lu cells and human skin fibroblasts were seeded on six-well plates as described (see MATERIALS AND METHODS) and cultured for 24 h. After the medium was changed to 2 ml of fresh serum-free DMEM, the cells were cultured for 16 h without (control) or with 10 µM SLPI or 5 ng/ml IL-1 in the absence (open columns) or presence (shaded columns) of 100 ng/ml IL-1ra. The amount of HGF released into the medium was determined by ELISA. Results are the means ± SEM for three different wells. Asterisks denote significant difference (P < 0.01) compared with the respective control cells.

Effect of Full-Length and C- and N-Terminal Domains of SLPI on HGF Production

Because the SLPI molecule is composed of two topologically superimposable structures (25), the N-terminal domain with antimicrobial activity (9) and the C-terminal domain with antiprotease activity, we next investigated the contribution of these peptide domains to HGF production and found that the C-terminal domain but not the N-terminal domain induced HGF (Figure 4). In these studies, the media from treated and untreated control cultures were assayed by ELISA for cumulative HGF production by human lung fibroblast CCD-11Lu cells. Both full-length SLPI (10 µM) and its C-terminal domain (10 µM) caused significantly higher HGF production than that caused by the control cultures (Figure 4A).

In the experiment to examine the dose dependency, full-length SLPI and its C-terminal domain (ranging from 0 to 10 µM) increased HGF production, although the latter showed about half the stimulation compared with that of the full-length SLPI (Figure 4B). These results show that the C-terminal domain but not the N-terminal domain of SLPI is effective in stimulating lung fibroblasts to produce HGF. The rate of HGF accumulation between 12 and 16 h was much higher than that at 12 h. When cycloheximide at 20 µg/ml was used, the stimulatory effect of SLPI was completely prevented (data not shown), indicating that the induction of HGF production in response to SLPI was dependent on de novo protein synthesis rather than on the simple release of the stored intracellular protein into the medium.

SLPI Stimulates HGF Production via a Pathway That Involves the Inhibition of Protease(s)

Because the HGF-inducing effect of SLPI could be attributed only to its C-terminal domain with antiprotease activity, we investigated the effects of protease inhibitors and found that ₁-AT also stimulated HGF production (Figure 5). In this context, CCD-11Lu cells were treated with 50 µg/ml human ₁-AT or 10 µM ONO-5046, as well as 10 µM full-length SLPI or 10 µM C-terminal domain of SLPI in serum-free medium. Stimulation with ₁-AT but not with a synthetic neutrophil elastase inhibitor (ONO-5046) showed a significant increase (2.9-fold) in HGF production, which was a little smaller than that by full-length SLPI (4.2-fold) but similar to that with the C-terminal domain (3.1-fold). In contrast, neither the N-terminal domain of SLPI with antimicrobial activity nor defensin, a multidisulfide peptide, had a significant effect on HGF production. These results suggest that the stimulation of HGF production in lung fibroblasts likely involves the inhibitor proteins for proteases.

View larger version (28K): [in this window] [in a new window]   Figure 5.   Responses of human lung fibroblasts to antiprotease and antibacterial agents. Human lung fibroblast CCD-11Lu cells were seeded on six-well plates as described in MATERIALS AND METHODS and cultured for 24 h. Subsequently, cells were incubated with 2 ml of serum-free DMEM for 16 h under the following conditions: no addition (control); 10 µM full-length SLPI (SLPI); 10 µM C-terminal domain of SLPI (C-SLPI); 50 µg/ml human 1-AT; 10 µM ONO-5046 (ONO-5046); 10 µM N-terminal domain of SLPI (N-SLPI); or 2.5 µg/ml human defensin HNP-1 (Defensin). The concentrations of HGF in the culture medium were measured by ELISA. Results are the means ± SEM for three different wells. Asterisks denote significant difference (P < 0.01) compared with control cells.

Effect of Oxidation by NCS on HGF-Inducing Activity of SLPI

Because the oxidation of the P'1 component (the amino acid position 1' in the reactive sites of serine protease inhibitors) of the protease inhibitory residue, Met⁷³ of human SLPI, results in the loss of the serine protease-inhibitory function (24), full-length SLPI or its C-terminal domain was inactivated by oxidation with NCS (Figure 6). A constant amount of full-length SLPI or its C-terminal domain (0.64 mM) was incubated for 2 h in the presence or absence of increasing concentrations of NCS, and their HGF-inducing activities were assessed using CCD-11Lu cells. The stimulatory effects of SLPI and its C-terminal domain were significantly inhibited by the pretreatment with NCS (P < 0.01 at the concentration of 32 mM, which corresponds to the 50:1 molar ratio of NCS to inhibitor). In contrast, increasing the concentration of NCS had no effect on the basal level of HGF production in CCD11Lu cells. These results provide further evidence that the involvement of an intact P'1 residue is required for the induction of HGF.

View larger version (31K): [in this window] [in a new window]   Figure 6.   Effect of NCS oxidation on the HGF-inducing activity of full- and half-length SLPIs. Full-length SLPI (SLPI, 0.64 mM final) or C-terminal domain of SLPI (C-SLPI, 0.64 mM final) was oxidized in 100 mM Tris/HCl, pH 8.3 without (none, open columns) or with 6.4 (striped columns), 12.8 (gray columns), or 32 mM NCS (filled columns), in a final volume of 62.3 µl. After 2 h at 25°C, the oxidation process was stopped by diluting the reaction with 51 µl of 100 mM N-acetylmethionine, and then the HGF-inducing activity was assayed. At 24 h before the assay, human lung fibroblast CCD-11Lu cells were seeded on 12-well plates as described (see MATERIALS AND METHODS) and cultured. Cells were incubated for another 16 h with 0.67 ml of serum-free medium containing oxidized full-length SLPI or C-terminal domain of SLPI at a final concentration of 10 µM. As a control, the equivalent of the NCS reaction solutions alone after dilution with N-acetylmethionine was used. The concentrations of HGF in the culture medium were determined by ELISA. Results are the means ± SEM for three different wells. Asterisks denote significant difference (P < 0.01) compared with the respective reaction without NCS (none).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

By protein purification and cDNA cloning, SF/HGF-inducing factor (SF-IF), which was detected in the conditioned media of a ras-transformed mouse 3T3 fibroblast cell line, has been shown to be a mouse homologue of human SLPI (15, 16). Because the cell types with SF-IF production and SF-IF response range widely across species barriers (15), we were prompted to examine whether recombinant human SLPI also has HGF-inducing activity. Our current study clearly demonstrates that recombinant human SLPI induced HGF production in human lung fibroblasts, and that this activity of HGF induction resides in the C-terminal domain of SLPI. Our study also shows that HGF induction by SLPI requires an intact active site for protease inhibition in this domain.

Mechanism of HGF Induction by SLPI

Treatment of human lung fibroblasts with SLPI augmented the expression of HGF mRNA. The inflammatory cytokine IL-1 also upregulates the HGF gene expression (21), and this stimulatory signal on human lung fibroblasts was confirmed to be transmitted through the IL-1 receptor in the present study by the fact that IL-1ra blocked the effect of IL-1. In marked contrast, IL-1ra failed to inhibit the SLPI-induced HGF production in human lung fibroblasts. These observations indicate that the signal evoked by SLPI is transmitted through a different route from that of the IL-1 receptor. Tamura and coworkers reported that tumor necrosis factor (TNF)- induced HGF in lung fibroblasts (28), and Matsumoto and colleagues showed that PGs (E₁, E₂, and I₂ analogue) also induced HGF in skin fibroblasts (29). It is of interest whether these novel cascades of HGF induction share common parts of the pathway.

The molecular mechanisms by which SLPI acts at the first step to evoke HGF induction in cultured fibroblasts required the protease-inhibitory activity of SLPI. This is supported by the following findings: (1) the truncated C-terminal domain of the SLPI molecule, in which the active site for the protease inhibition (Lue⁷²-Met⁷³) is included, had the ability to induce HGF despite the absence of the N-terminal domain; (2) oxidative inactivation of the protease inhibitory activity of the C-terminal domain of SLPI abolished the HGF-inducing activity of SLPI; and (3) ₁-AT, another serine protease inhibitor, also induced HGF production by human lung fibroblasts. However, the inhibition of serine proteases per se does not seem to be sufficient to induce HGF because the synthetic and competitive elastase inhibitor ONO-5046 (30) failed to induce HGF production. This discrepancy may be explained by a target protease that is complexed with ₁-AT or SLPI but not by ONO-5046.

Although no serine protease involved in HGF production has been known to date, the report demonstrating that HGF increases the expression of urokinase-type plasminogen activator (u-PA), a serine protease (31), suggests that an unidentified serine protease whose production from fibroblasts is induced by HGF may in turn participate in an autocrine feedback mechanism and negatively modulate the quantity of the HGF production. In this context, not a few serine proteases have been reported to be secreted from fibroblasts, including u-PA (32), plasmin (33), fibroblast-activation protein (34), a 92-kD protease cleaving insulin-like growth factor binding protein-5 (35), tissue-type plasminogen activator (36), a 43-kD target protease of the Bowman-Birk protease inhibitor (37), and a calcium-dependent serine protease (38). Thus, one possible picture that can explain the mechanism of the SLPI- induced HGF production by fibroblasts is that SLPI interacts and inactivates an autocrine serine protease that is functioning to downregulate the HGF synthesis of fibroblasts presumably near the surface of cells, then stimulating the HGF production. Alternatively, an unknown domain of SLPI protein may interact with the surface receptor on the lung fibroblasts that can signal to the HGF upregulation. To identify the target protein of SLPI, two hybrid experiments could be useful for further elucidation.

Mechanism by which Serine Protease Inhibitors Ameliorate Lung Injury

Although protease inhibitors are expected to target specific proteolytic reactions and hence to protect substrates from the destructive effects of proteases, their additional functions may affect the total anti-inflammatory capability. It is of interest that intraperitoneal administration of ₁-AT or the truncated C-terminal domain of human SLPI has been shown to ameliorate BLM-induced lung injury and the subsequent fibrosis in vivo (13, 39). However, the elevated level of the elastase activity in the lung after BLM treatment was not reduced by this exogenous supplementation of ₁-AT or SLPI. Further, intraperitoneal administration of ₁-AT affected neither the superoxide production nor chemotactic activity for neutrophils in bronchoalveolar lavage fluid (39). In this context, the local induction of HGF by SLPI or ₁-AT may contribute to the amelioration of lung injury because exogenous recombinant human HGF suppressed BLM-induced lung injury and fibrosis in vivo in our previous study (18).

How serine protease inhibitors evoke the gene expression remains unclear. The fact that oxidation of the methionine residue in the active center of SLPI resulted in the loss of the HGF-inducing activity suggests the importance of the interaction and complex formation between proteases and inhibitors. It is of interest that Perlmutter and associates reported the existence of serpin-enzyme complex receptor in several cell lines (40). Although there was no definite difference in the effects on HGF mRNA expression of human lung fibroblasts between the SLPI alone and the neutrophil elastase-SLPI complex (data not shown), further study may predict the existence of such receptors; a protease-inhibitor molecule not only functions as an inhibitor of the target protease but also exerts a complicated effect of anti-inflammation via serpin-enzyme complex in signaling protective gene expression such as HGF, as in the current study. Additionally, cell-type specificity of such receptors may account for the different responses to SLPI treatment between lung and skin fibroblasts.

SLPI Regulates the Gross Anti-Inflammatory Reaction

In addition to the inhibition of serine proteases and the induction of HGF, SLPI has recently been shown to have diverse effects in vitro that are thought to counteract the progression of inflammation in vivo: (1) suppression of COX-2, resulting in the reduction of PGE₂ and metalloproteinases in monocytes (11); (2) the suppression of HIV-I infection to lymphocytes (9, 10); and (3) antimicrobial activities similar to those of defensins (8). It is of interest that SLPI also counteracts the LPS-induced stimuli in macrophages (6). In this context, the SLPI cDNA has been cloned by differential display as one of the genes specifically expressed in the macrophages from an LPS-hyporesponsive strain of mice. Ectopic expression of the SLPI cDNA in macrophages from the LPS-responsive strain shut down the signal transduction pathways elicited by LPS, such as the stimulation of TNF- production, nitric oxide synthesis, and nuclear factor-B binding activity of the nuclear protein (6). In addition, IL-10 or IL-6 induced SLPI in macrophages in vivo (7). These findings strongly suggest the physiologic function of SLPI as a mediator of anti-inflammatory reactions in the cell, making it one of the possible candidates for anti-inflammatory therapeutics.

In conclusion, the present study provides the first evidence that SLPI increases HGF production in human lung fibroblasts by regulating the steady-state mRNA level. Further, an intact active site for protease inhibition is indispensable for this stimulatory effect. We hypothesize that SLPI as a serine protease inhibitor may coordinate the overall anti-inflammatory effect not only by inhibiting the proteases derived from inflammatory cells, but also by ameliorating inflamed tissue at least in organs such as lung by upregulating the expression of anti-inflammatory and regenerative genes such as HGF and downregulating the inflammatory genes.