Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Sequestration of peripheral blood neutrophils and monocytes within the lung is characteristic of a number of acute and chronic pulmonary diseases (1). The presence of neutrophils is reflected by the local generation of chemotactic agents, which direct neutrophil migration from the vascular compartment into the alveolar space along chemotactic gradients. Additionally, alveolar macrophages (AM) are derived predominantly from differentiated peripheral blood monocytes, and to a limited extent from local macrophage replication (6). Although chemotactically activated neutrophils and macrophages serve a vital role in host defense against a number of organisms, the presence of increased numbers of activated neutrophils and macrophages can lead to tissue injury through the excessive elaboration of inflammatory cytokines, proteolytic enzymes, and oxygen radicals (2, 9). Substantial investigation has focused on AM as a primary source of chemotactic factors (10). However, neutrophil chemotactic activity (NCA) and monocyte chemotactic activity (MCA) have been found to be produced by endothelial cells (13), fibroblasts (14), and pulmonary epithelial cells (15).

The fibroblast is the principal cell of most connective tissues, and is involved in producing collagenous and noncollagenous components of the extracellular matrix. This synthetic activity serves an important structural function by providing a framework for organ integrity. Recent studies have shown that in addition to maintaining connective tissue, fibroblasts are also important participants in the orchestration of acute and chronic inflammation. In this context, fibroblasts release interleukin (IL)-8, monocyte chemoattractant protein-1 (MCP-1), granulocyte-macrophage colony-stimulating factor (GM-CSF), and transforming growth factor (TGF)- in response to IL-1, tumor necrosis factor (TNF)-, and platelet-derived growth factor, suggesting that they contribute to certain disease states (18).

Activation of the kallikrein-kinin system in lung injury has long been recognized. Bradykinin (BK) is generated from kininogens by the actions of plasma and tissue kallikreins (kininogenases) (26, 27). The effects of BK on pulmonary circulation and lung mechanics have been intensively evaluated. BK also stimulates AM and airway epithelial cells to release chemotactic factors for inflammatory cells (28, 29). Recently, a bradykinin B₂ (BKB₂) receptor antagonist was found to attenuate the acute lung injury induced by infusion of live Pseudomonas aeruginosa, including the migration of neutrophils into the lung and their sequestration in the lung (30). In this context, BK may participate in the release of inflammatory mediators from lung cells.

Although airway epithelial cells and AM may play a role in inflammatory cell migration from the interstitium to the alveolar and bronchial spaces in response to BK (28, 29), the mechanism underlying inflammatory cell migration from the vascular compartment to the interstitium remains to be elucidated. The role of human lung fibroblasts (HLF) in inflammatory cell recruitment from the vascular compartment to the interstitium in response to BK is uncertain. The purpose of the present study was to determine whether lung fibroblasts could participate in the recruitment of inflammatory cells to the lung interstitium. Specifically, we evaluated the possibility that lung fibroblasts release NCA and MCA in response to BK. The results showed that HLF can release NCA and MCA in response to BK, including leukotriene B₄ (LTB₄), IL-8, granulocyte colony-stimulating factor (G-CSF), MCP-1, GM-CSF, and TGF-.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Culture of Human Lung Fibroblasts

We used fetal HLF (lung, diploid, human, passage 27) from an established, commercially available cell line (American Type Tissue Culture Collection, Rockville, MD). HLF were suspended at 1.0 × 10⁶ cells/ml in F-12 medium (GIBCO, Grand Island, NY) supplemented with penicillin (50 U/ ml; GIBCO), streptomycin (50 µg/ml; GIBCO), fungizone (2 µg/ml; GIBCO), and 10% fetal calf serum (GIBCO). HLF suspensions (3 ml) were added to 30-mm-diameter tissue culture dishes (Corning, Corning, NY) and were cultured at 37°C in a 5% CO₂ atmosphere. After 4-6 d in culture, the cells had reached confluence, and the culture medium was then replaced with 2 ml of medium supplemented as described earlier and the cells were incubated for 1 d more.

Measurement of BK Concentrations in HLF Cultures

We measured the concentrations of BK at selected time points (12, 24, 48, and 72 h) to evaluate the half-life and duration of the stimulating potential of BK after exposure of fibroblasts to HLF. BK was measured with a highly sensitive radioimmunoassay (RIA) as previously described (30).

Exposure of HLF to BK

Medium was removed from cells by washing twice with serum-free F-12, and cells were then incubated in the presence and absence of BK (Sigma, St. Louis, MO). To determine the dose- and time-dependent release of NCA and MCA, the cultures were incubated at various concentrations of BK (0.01, 0.1, 1.0, 10, and 100 µM) for 12, 24, 48, 72, and 96 h at 37°C in a humidified 5% CO₂ atmosphere. BK at 100 µM did not cause HLF injury (no deformity of cell shape or detachment from the tissue culture dish, and viability of more than 95% of cells by trypan blue dye exclusion) after 96 h of incubation. The supernatant fluids were then harvested and stored at 80°C until assayed. At least six separate HLF supernatant fluids were harvested for each experimental condition.

Measurement of NCA and MCA

Polymorphonuclear leukocytes were purified from heparinized normal human blood by the method of Boyum (31). Briefly, 15 ml of venous blood was obtained from healthy volunteers and was then sedimented with 3% dextran in isotonic saline for 45 min in order to separate the white blood cells from red blood cells. The leukocyte-rich upper layer was collected, and neutrophils were separated from mononuclear cells by Ficoll-Hypaque density centrifugation (Histopaque 1077; Sigma). Contaminating red blood cells were removed by using a lysing solution consisting of 0.1% KHCO₃ and 0.83% NH₄Cl. The suspension was then centrifuged at 400 × g for 5 min and washed three times in Hanks' balanced salt solution (HBSS; Biofluids, Rockville, MD). The resulting cell pellet, as determined by trypan blue dye and erythrosin exclusion, consisted of > 96% neutrophils and > 98% viable cells. The cells were suspended in Gey's balanced salt solution (GBSS; GIBCO) containing 2% bovine serum albumin (BSA; Sigma) at pH 7.2 to give a final concentration of 3.0 × 10⁶ cells/ml. This suspension was used for the neutrophil chemotaxis assay.

Mononuclear cells for the chemotaxis assay were obtained from normal human volunteers by Ficoll-Hypaque density centrifugation to separate red blood cells and neutrophils from mononuclear cells. The mononuclear cells were harvested at the interface. The suspension was then centrifuged at 400 × g for 10 min and washed three times in HBSS. The resulting preparation routinely consisted of 30% large monocytes and 70% small lymphocytes as determined morphologically and by -naphthyl acetate esterase staining (Sigma), with > 98% viability as assessed by trypan blue dye and erythrosin exclusion. The cells were suspended in GBSS containing 2% BSA at pH 7.2 to give a final concentration of 5.0 × 10⁶ cells/ml. This suspension was then used for the monocyte chemotaxis assay.

The chemotaxis assay was performed in a 48-well microchemotaxis chamber (NeuroProbe Inc., Cabin John, MD) as previously described (32). The bottom wells of the chamber were filled with 25 µl of fluid containing the chemotactic stimulus or medium in duplicate. A 10-µm- thick polyvinylpyrrolidone-free polycarbonate filter (Nucleopore, Pleasanton, CA), with a pore size of 3 µm for neutrophil chemotaxis and 5 µm for monocyte chemotaxis, was placed over the bottom wells. The silicon gasket and upper pieces of the chamber were applied, and 50 µl of the cell suspension was placed into the upper wells above the filter. The chambers were incubated in humidified air in 5% CO₂ at 37°C for 30 min for neutrophil chemotaxis and for 90 min for monocyte chemotaxis. After incubation, the chamber was disassembled and cells that had not migrated were wiped away from the filter. The filter was then immersed in methanol for 5 min, stained with Diff-Quik (American Scientific Products, McGaw Park, IL), and mounted on a glass slide. Cells that had completely migrated through the filter were counted in 10 random high-power fields (hpf, magnification ×1,000) per well by light microscopy.

To ensure that monocytes and not lymphocytes were the primary cells that migrated in the monocyte chemotaxis assay, some membranes were stained with -naphthyl acetate esterase according to the manufacturer's directions (Sigma).

To determine whether migration was due to movement along a concentration gradient (chemotaxis) or to stimulation of random migration (chemokinesis), a checkerboard analysis was done with HLF supernatant fluids harvested at 72 h in response to 100 µM BK (33). To do this, various dilutions of HLF supernatant fluids (1:1, 1:4, 1:16, 1:64, 1:256) were placed below the membrane of the microchemotaxis chamber and the target cells were placed above the membrane.

Molecular Sieve Column Chromatography of Chemotactic Activity

To determine the approximate molecular weight of the released chemotactic activity in the supernatant fluids harvested at 72 h in response to 100 µM BK, we performed molecular sieve column chromatography, using Sephadex G-100 (Pharmacia, Piscataway, NJ). The supernatant fluid was not concentrated for passage through the column. The column size was 10 mm × 900 mm. The bet volume was almost 60 ml. We put 2 ml of supernatant fluid into the column. HLF culture supernatant fluid was eluted with phosphate-buffered saline at a flow rate of 6 ml/h, and fractions after the void volume were evaluated in duplicate for NCA and MCA. The chemotactic activity might have been diluted by column chromatography; however, we collected 1 ml per fraction. Because peaks usually consisted of five to seven fractions (i.e., 5 to 7 ml), the chemotactic activity of each molecular-weight peak was not extensively diluted.

Effects of Metabolic Inhibitors on the Release of NCA and MCA

The effects of the nonspecific lipoxygenase inhibitors nordihydroguaiaretic acid (NDGA; 100 µM; Sigma) and diethylcarbamazine (DEC; 1 mM; Sigma), and of the 5-lipoxygenase inhibitor AA-861 (100 µM; Takeda Pharmaceutical Co., Tokyo, Japan) on the release of NCA and MCA were evaluated in response to 100 µM BK after 72 h incubation. To further examine the involvement of protein synthesis in the release of chemotactic activity, we added cycloheximide (10 µg/ml; Sigma) to inhibit protein synthesis (34). At the concentrations used, NDGA, DEC, and AA-861 inhibited the release of LTB₄ in HLF cultures, and did not cause cytotoxicity to HLF after 72 h incubation.

Identification of Target BK Receptor on HLF

To determine the receptor responsible for the release of chemotactic activity, the BKB₁ receptor antagonist Des-Arg⁹-[Leu⁸]-BK (Sigma) and the BKB₂ receptor antagonist D-Arg-[Hyp³Thi^5,8-d-Phe⁷]-BK at concentrations of 100 µM each were used. Thirty minutes before the addition of 50 µM BK, HLF were treated with BKB₁ and BKB₂ receptor antagonists and incubated for 72 h.

Effects of LTB₄ and Platelet-Activating Factor Receptor Antagonists on NCA and MCA

Because the release of both NCA and MCA was blocked by 5-lipoxygenase inhibitors, and because both NCA and MCA were extracted into ethyl acetate, an LTB₄ receptor antagonist (ONO 4057; ONO Pharmaceutical Co., Tokyo, Japan) and a platelet-activating factor (PAF) receptor antagonist (TCV-409; Takeda), at a concentration of 10⁵ M in each case, were used to evaluate the involvement of LTB₄ in NCA and MCA in the crude supernatant fluids and in the fraction of the lowest molecular mass separated by column chromatography (35, 36).

Measurement of LTB₄ and PAF in Supernatant Fluids

The concentration of LTB₄ in the supernatant fluids was measured with an RIA as previously described (37). Anti-LTB₄ serum, [5, 6, 8, 9, 11, 12, 14, 15, ³H (N)]-LTB₄, and synthetic LTB₄ were purchased from Amersham Inc. (Arlington Heights, IL). Briefly, ethanol-and-supernatant mixtures were centrifuged at 5,500 × g at 0°C. The supernatants were then evaporated under N₂ gas at 37°C to remove the ethanol. To each sample, 10 ml of distilled water was added. The diluted samples were then acidified to pH 4.0 with 0.1 N HCl and were applied to Sep-Pak C18 columns (Waters Associates, Milford, MA). The columns were washed with a mixture of 10 ml of distilled water and 20 ml of petroleum ether, and were then eluted with 15 ml of methanol. The eluates were dried with N₂ gas at 37°C and then redissolved in 20 µl of methanol and 180 µl of RIA buffer (50 mM Tris-HCl buffer containing 0.1% [wt/vol] gelatin, pH 8.6). [³H]LTB₄ was diluted in RIA buffer (0.1 ml, containing approximately 4,000 dpm) and mixed with 0.1 ml of standards or samples in disposable siliconized tubes. Anti-LTB₄ serum, diluted with RIA buffer (0.1 ml), was added to the siliconized tubes to give a total incubation volume of 0.4 ml. The resulting mixture was incubated at 4°C for 18 h. Free LTB₄ was absorbed onto dextran-coated charcoal. Following centrifugation for 15 min at 2,000 × g, the supernatant containing the antibody-bound LTB₄ was decanted into the well of a scintillation counter (Tricarb-3255; Packard Co., Downers Grove, IL). Scintillation fluid (Aquazol 2; NEN Co., Boston, MA) was added, and radioactivity was counted for 4 min.

PAF in the supernatant fluids was quantitated with the scintillation proximity assay system. Briefly, this assay system combines the use of a tritiated PAF tracer of high specific activity with an antibody specific for PAF and a PAF standard similar to that used in the method for measuring LTB₄.

Effects of Polyclonal Antibodies to IL-8, G-CSF, MCP-1, GM-CSF, TGF-, and Regulated on Activation, Normal T-cell Expressed and Secreted

Neutralizing antibodies to IL-8, G-CSF, MCP-1, GM-CSF, TGF-, and regulated on activation, normal T-cell expressed and secreted (RANTES) (Genzyme, Cambridge, MA) were added to the HLF supernatant fluids that were harvested at 72 h after exposure to 100 µM of BK at the concentrations recommended for inhibiting these cytokines, and the resulting preparations were incubated for 30 min at 37°C. The antibodies to IL-8, G-CSF, MCP-1, GM-CSF, TGF-, and RANTES were tested by blockade of the chemotactic response of neutrophils and monocytes to each human recombinant cytokine. The HLF samples containing the antibodies were then used for chemotactic assay. The antibodies did not influence the chemotactic response to endotoxin-activated serum (data not shown). To assess the nonspecific effect of immunoglobulin (Ig)G, nonimmune IgG (Genzyme) was added to the supernatant fluids, incubated for 30 min at 37°C, and used for chemotactic assay.

Measurement of IL-8, G-CSF, MCP-1, GM-CSF, TGF-, and RANTES

The concentrations of IL-8, G-CSF, MCP-1, GM-CSF, TGF-, and RANTES in HLF supernatant fluids cultured for 72 h at the BK concentration of 100 µM were measured with enzyme linked immunosorbent assays (ELISAs) according to the manufacturers' directions. Assay kits for GM-CSF and RANTES were purchased from Amersham (Buckinghamshire, UK), and the minimum concentrations detected with these assays were 2.0 pg/ml for GM-CSF, and 15.6 pg/ml for RANTES. IL-8, MCP-1, and TGF- kits were purchased from R&D Systems (Minneapolis, MN), and the minimum detectable concentration of IL-8, MCP-1, and TGF- was 10.0 pg/ml, 31.3 pg/ml, and 0.31 ng/ml, respectively. A chemiluminescence enzyme immunoassay (CLEIA) assay kit for G-CSF was obtained from Chugai Pharmaceutical Co. (Tokyo, Japan), with a minimum detectable G-CSF concentration of 1.0 pg/ml.

Statistics

In experiments in which multiple measurements were made, differences between groups were tested for significance through one-way analysis of variance, with Fisher's multiple-range test applied to data at specific times and dose points. In experiments in which single measurements were made, differences between groups were tested for significance through Student's paired t test. In all cases, a value of P < 0.05 was considered significant. Data in figures and tables are expressed as mean ± SE.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Concentrations of BK in HLF Cultures

BK at the initial concentration of 50 µg/ml declined to 1.32 ± 0.01 µg/ml at 12 h, to 0.00523 ± 0.000112 µg/ml at 24 h, and to 0.000101 ± 0.000009 µg/ml at 48 h, becoming undetectable at 72 h (n = 3 monolayers). Thus, the half-life of BK was less than 12 h, and very litle BK was present in the supernatant fluid after 72 h. Because most of the BK in the supernatant fluids was metabolized by 72 h, the exposure of HLF to BK was observed after 96 h.

Release of NCA and MCA from HLF

HLF released NCA and MCA in a dose-dependent manner in response to BK (Figures 1A and 1B). The lowest doses of BK that stimulated HLF were 0.1 µM for neutrophils and 0.01 µg/ml for monocytes. Increasing concentrations of BK, to 100 µM, progressively increased the release of NCA and MCA. Although HLF released NCA and MCA constitutively, HLF further released NCA and MCA in response to BK in a time-dependent manner (Figures 2A and 2B). The release of NCA and MCA was significant after 24 h exposure to BK (Figures 2A and 2B). The release of chemotactic activity reached a plateau at 72 h. BK itself did not show any chemotactic activity for either neutrophils or monocytes (data not shown). The chemotactic responses to LTB₄ and to formyl-methionyl-leucyl-phenylalanine (FMLP) as a positive control at a concentration of 10⁷ M for each were 102.0 ± 7.4 and 118.5 ± 8.9 cells/hpf, respectively, for neutrophils, and 75.6 ± 3.4 and 84.3 ± 7.8 cells/hpf, respectively, for monocytes. The chemotactic responses to normal saline as a negative control were 10.5 ± 2.7 cells/hpf for neutrophils and 7.3 ± 1.8 cells/hpf for monocytes.

View larger version (15K): [in this window] [in a new window]   Figure 1.   Dose-dependent release of NCA (A) and MCA (B) from HLF monolayers in response to BK after 72 h incubation (n = 8). Values are expressed as means ± SE. * P < 0.05 versus without bradykinin; ** P < 0.01 versus without bradykinin; *** P < 0.001 versus without bradykinin.

View larger version (16K): [in this window] [in a new window]   Figure 2.   Time-related release of NCA (A) and MCA (B) in response to 100 µM BK, and baseline release of NCA and MCA from HLF monolayers (n = 8). Closed circles represent presence of BK and open circles represent absence of BK. Values are expressed as means ± SE. * P < 0.05 compared with medium alone; ** P < 0.01 compared with medium alone; *** P < 0.001 compared with medium alone; #P < 0.05 compared with supernatant fluids without BK; ##P < 0.01 compared with supernatant fluids without BK.

Checkerboard analysis revealed that the HLF supernatant fluids to which BK was added induced neutrophil migration in the presence of a gradient across the membrane in a concentration-dependent manner (data not shown). The passage of neutrophils was therefore consistent with chemotactic rather than with chemokinetic migration. In contrast, monocyte migration was induced only slightly in the absence of a gradient (data not shown), indicating that it was predominantly chemotactic and partly chemokinetic.

Confirmation that the cells that had migrated were monocytes was provided by the findings that: (1) > 90% of the cells that migrated appeared to be monocytes morphologically by light microscopy; (2) > 90% of the cells that migrated were esterase positive; and (3) lymphocytes purified by allowing the monocytes to attach to plastic, and tested in the chemotaxis assay, showed only 0-20% of the chemotactic activity of the monocyte preparation.

Inhibition of the Release of Chemotactic Activity by Metabolic Inhibitors

The supernatant fluids incubated with 100 µM BK in the presence of NDGA, DEC, and AA-861 showed a decrease in the release of NCA and MCA (P < 0.001; Figures 3A and 3B). Cycloheximide inhibited the release of both NCA (P < 0.001; Figure 3A) and MCA (P < 0.001; Figure 3B).

View larger version (21K): [in this window] [in a new window]   Figure 3.   Inhibition of release of NCA (A) and MCA (B) in response to 100 µM BK with 72 h incubation by DEC, NDGA, AA-861, and cycloheximide (n = 8). Chemotactic activity is the ordinate and experimental groups are on the abscissa. Values are expressed as means ± SE. * P < 0.001 compared with untreated supernatant fluids.

Molecular Sieve Column-Chromatographic Findings for Released Chemotactic Activities

Molecular sieve column chromatography with Sephadex G-100 revealed that NCA was heterogeneous in size (Figure 4A). At least three peaks of NCA were separated by column chromatography, with estimated molecular weights above and below that of cytochrome C (MW: 12,300 D) and an additional peak that eluted after quinacrine (MW: 450 D). The released MCA was also heterogeneous (Figure 4B). At least four peaks of MCA appeared to be separated by column chromatography, with estimated molecular weights below those of BSA and cytochrome C, and an additional peak that eluted after quinacrine.

View larger version (19K): [in this window] [in a new window]   Figure 4.   Molecular sieve column chromatographic findings for NCA (A) and MCA (B) released into HLF supernatant fluids in reponse to 100 µM BK and harvested at 72 h. The data shown represent four experiments. NCA and MCA are on the ordinate and fraction number is on the abscissa. Closed squares represent presence of BK and closed circles represent absence of BK. Molecular weight (MW) markers BSA (66,000), cytochrome C (12,400), and quinacrine (473) are indicated by arrows.

Inhibition of NCA and MCA by LTB₄ Receptor Antagonists

Both the NCA and MCA of crude samples were significantly inhibited by addition of the LTB₄ receptor antagonist ONO 4057, by about 70% for NCA and 40% for MCA (P < 0.001; Figures 5A and 5B). ONO 4057 also inhibited the column chromatographically separated lowest-molecular-weight peaks of NCA and MCA (by about 80% for NCA and 60% for MCA). In contrast, PAF receptor antagonist had no effects on NCA or MCA. LTB₄ and PAF receptor antagonists, each at a concentration of 10⁵ M, completely inhibited neutrophil migration in response to a 10⁷ M concentration of LTB₄ and PAF, but showed no inhibitory effects on FMLP- or endotoxin-activated serum-induced neutrophil and monocyte chemotaxis (data not shown).

View larger version (20K): [in this window] [in a new window]   Figure 5.   Inhibition by LTB4 receptor antagonist but not by PAF receptor antagonist of NCA (A) and MCA (B) in supernatant fluids induced by 100 µM BK and harvested at 72 h (n = 8). Values are expressed as means ± SE. * P < 0.001 compared with crude supernatant fluids.

Release of LTB₄ from HLF

The measurement of LTB₄ in supernatant fluids with RIA revealed that HLF released a significant quantity of LTB₄ in the baseline culture condition. However, the addition of BK at a concentration of 100 µM for 72 h did not induce LTB₄ release from HLF (244 ± 20 pg/ml [control] versus 254 ± 9 pg/ml [BK]; P = 0.10). PAF was not detected in the baseline or BK-stimulated supernatant fluids (i.e., was present at less than 40 pg/ml).

Inhibition of NCA and MCA by Polyclonal Antibodies to IL-8, G-CSF, MCP-1, GM-CSF, TGF-, and RANTES

Because HLF had the potential to release chemokines, and because chemokines released from HLF might be responsible for NCA and MCA, we used polyclonal blocking antibodies to IL-8, G-CSF, MCP-1, GM-CSF, TGF-, and RANTES to study release of these chemokines. Among these antibodies, antibodies to IL-8 and G-CSF inhibited NCA (P < 0.01; Figure 6A). Antibodies to MCP-1 and GM-CSF inhibited MCA (P < 0.01; Figure 6B). Anti-TGF- antibody slightly but significantly attenuated MCA (P < 0.05; Figure 6B). In contrast, antibody to RANTES did not inhibit MCA. We evaluated the effect of antibodies to IL-8, G-CSF, MCP-1, GM-CSF, and TGF- on column chromatographically separated NCA and MCA. These antibodies also inhibited NCA and MCA at molecular weights corresponding to the respective chemotactic peaks by about 60-80%.

View larger version (24K): [in this window] [in a new window]   Figure 6.   Effects of blocking antibodies to IL-8, G-CSF, MCP-1, GM-CSF, and TGF- on NCA (A) and MCA (B) released into HLF supernatant fluids in response to 100 µM BK with 72 h incubation (n = 8). Values are expressed as means ± SE. * P < 0.05 compared with crude supernatant fluids. ** P < 0.01 compared with crude supernatant fluids.

Release of IL-8, MCP-1, G-CSF, GM-CSF, TGF-, and RANTES from HLF by BK

Measurement of chemotactic cytokines through ELISA revealed that BK at a concentration of 100 µM, following 72 h incubation, stimulated the release of IL-8 and G-CSF as NCA (P < 0.001; Table 1) and of GM-CSF, MCP-1, and TGF- as MCA (P < 0.05; Table 1). In contrast, RANTES was not detected in HLF supernatant fluids.

Because NCA comprised IL-8 and G-CSF, and because MCA comprised MCP-1 and GM-CSF, we evaluated the individual time course of mediators that exhibited chemotactic activity (Figures 7A throught 7D). All chemokines showed time-dependent increases in concentration.

View larger version (25K): [in this window] [in a new window]   Figure 7.   Individual time courses of chemokines that comprised chemotactic activity. Release of IL-8 (A), G-CSF (B), MCP-1 (C), and GM-CSF (D) from HLF in response to 100 µM BK (n = 6).

Inhibition of Release of NCA, MCA, and Individual Chemokines by Both BKB₁ and BKB₂ Receptor Antagonists

Both BKB₁ and BKB₂ receptor antagonists significantly inhibited the release of NCA and MCA harvested in both cases after 72 h incubation (Figures 8A and 8B). The individual chemokines (i.e., IL-8, G-CSF, MCP-1, and GM-CSF) were significantly inhibited by BKB₁ and BKB₂ receptor antagonists (Figures 9A through 9D).

View larger version (14K): [in this window] [in a new window]   Figure 8.   Effects of BKB1 and BKB2 receptor antagonists (each at 100 µM) on the release of NCA (A) and MCA (B) from HLF in response to 50 µM of BK after 72 h incubation (n = 5). * P < 0.01 compared with BK-stimulated supernatant fluids.

View larger version (22K): [in this window] [in a new window]   Figure 9.   Effects of BKB1 and BKB2 receptor antagonists (each at 100 µM) on the release of IL-8 (A), G-CSF (B), MCP-1 (C), and GM-CSF (D) by HLF in response to 50 µM BK after 72 h incubation (n = 6). * P < 0.01 compared with BK-stimulated supernatant fluids.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

BK is thought to be a potent inflammatory mediator in lung disease (26, 27). Turino and coworkers suggested that immunoreactive kinins and kininogenase activity were present in bronchoalveolar lavage fluid and plasma obtained from patients with lung inflammation (27). BK is generated from kininogens by the action of plasma and tissue kallikreins (26, 27), and its action includes vasodilation, vascular leakage, and contraction of smooth muscle (24, 27). BK has also been found to stimulate alveolar and bronchial epithelial cells and AM to release NCA and MCA (28, 29). In the present study, BK induced the release of NCA and MCA from HLF, which may contribute to inflammatory cell extravasation from the vasculature to the interstitium.

Although the potential of HLF to release IL-8, G-CSF, GM-CSF, MCP-1, and TGF- in response to BK was less than that observed upon stimulation of HLF with TNF- and IL-1 (data not shown), the release of NCA and MCA was increased by 3- to 4-fold compared with the constitutive release of chemotactic activity. The chemotactic potential of the activity released from HLF was greater than that released from 10⁶ AM per culture dish in response to endotoxin (unpublished data). It is therefore likely that BK contributes to the recruitment of inflammatory cells to the lung interstitium from the vasculature by stimulating HLF.

It has been reported that lung fibroblasts have the potential to release IL-8 and G-CSF in response to TNF- or IL-1 (38, 39). In the present study, blocking antibodies to IL-8 and G-CSF attenuated NCA to a similar level of chemotactic activity. BK stimulated the release of IL-8 and G-CSF from HLF. Although IL-8 is well known as a neutrophil chemotactic factor in a variety of inflammatory diseases, the role of G-CSF in exerting neutrophil chemotactic activity is less certain. In the present study, HLF released G-CSF as NCA in response to both BK and IL-8.

Wang and colleagues (40) reported that the concentration of G-CSF as NCA was 7-70 ng/ml. However, the concentration of G-CSF detected in the HLF supernatant fluids in our study was less than that reported (40). We produced neutrophil chemotaxis by using human recombinant G-CSF. The chemotactic concentration of G-CSF as NCA ranged from 10 to 100 pg/ml (unpublished data). The variation in the G-CSF concentration as a neutrophil chemotactic factor may have been due to differences in neutrophil separation. Because the blocking antibodies to G-CSF that were used in our study inhibited both total NCA in the supernatant fluids and NCA in the column chromatographically separated peaks, the contribution of G-CSF to NCA may be as a direct chemoattractant rather than through the activation of neutrophils.

The identification of NCA and MCA released from HLF is not complete. However, the inhibition of release of NCA and MCA by cycloheximide treatment suggests that these activities are at least partly dependent on protein synthesis (41). NCA and MCA were attenuated by antibodies to IL-8, G-CSF, MCP-1, GM-CSF, and TGF-. The concentrations of IL-8, G-CSF, MCP-1, GM-CSF, and TGF- in the supernatant fluids reached reported concentrations of NCA and MCA (42). Accordingly, HLF at least partly released IL-8, G-CSF, MCP-1, GM-CSF, and TGF- in the form of NCA and MCA.

HLF have the potential to release MCP-1, GM-CSF, and TGF-. However, the predominant MCA found in our study was in the form of GM-CSF and MCP-1, rather than TGF-. The release of GM-CSF from HLF in response to BK was striking compared with its release from alveolar or bronchial epithelial cells (data not shown). Because GM-CSF is one of glycoproteins that can stimulate the in vitro proliferation and differentiation of macrophage progenitor cells (45, 46), its augmented release from fibroblasts in response to BK suggest that fibroblasts rather than airway epithelial cells may be profoundly involved in macrophage recruitment, differentiation, and proliferation, and in macrophage activation in the lung interstitium.

The inhibition of release of NCA and MCA by NDGA, DEC, and AA-861 suggests that these activities are composed of lipoxygenase products. Both NCA and MCA were inhibited by LTB₄ receptor antagonist. Although the release of LTB₄ from HLF in response to BK was not significant compared with the control value, the concentration of LTB₄ reached the chemotactic range for neutrophils and monocytes. Thus, LTB₄ released constitutively from fibroblasts may be one of the important chemoattractants for neutrophils and monocytes (47).

BK has been reported to have several potent effects on airway functions, many of which are thought to be mediated by the BKB₂ receptor. The existence of BKB₂ receptors has been mapped on human and guinea pig lungs by autoradiography with [³H]BK (48). BKB₁ receptor ligands do not block [³H]BK binding in trachea and lung, indicating the absence of BKB₁ receptors and the presence of BKB₂ receptors in the healthy lung (49). However, the BKB₁ receptor antagonist Des Arg⁹-[Leu⁸]-BK was found to significantly inhibit airway hyperresponsiveness and neutrophilia induced by antigen challenge (50). Human lung fibroblasts of the WI-38 cell line express BKB₁ and BKB₂ receptors (39), and BK stimulates NF-kB activation and also induces IL-1 mRNA expression via the BKB₂ receptor. IL-1 production by BK-stimulated cells may also exert a positive effect on the expression of both BKB₁ and BKB₂ receptors. IL-1 is known to be one of the most potent inducing agents for the BKB₁ receptor (51, 52), and it may also increase the expression of functional BKB₂ receptors (53). Collectively, these findings suggest that both BKB₁ and BKB₂ receptors were involved in the release of NCA and MCA and of individual chemokines (i.e., IL-8, G-CSF, GM-CSF, and MCP-1) in the present study.

Although TGF- was detected in HLF supernatant fluid, antibody to TGF- did not attenuate MCA in our study. TGF- induces monocyte chemotaxis at concentrations of 0.1-10 pg/ml (43). At higher concentrations of TGF-, the chemotactic response of monocytes declines. It has been reported that the biologically inactive form of TGF-, which constitutes more than 98% of autocrine TGF-, is secreted by 12 different cell types (56). It was also found that TGF- was unable to bind to its receptor without prior proteolytic activation (56). The release of inactive TGF- may account for the failure of anti-TGF- antibody to inhibit MCA in the HLF supernatant fluids in our study.

In conclusion, we found that BK stimulated HLF to release NCA and MCA. The NCA and MCA released in response to BK included IL-8, G-CSF, MCP-1, GM-CSF, TGF-, and LTB₄. These results suggest that BK may play a role in inflammatory cell recruitment from the vasculature to the interstitium by stimulating the release of chemotactic activity from lung fibroblasts.