Neutrophil elastase induces mucin production by ligand-dependent epidermal growth factor receptor activation

Kazuhiro Kohri, Iris F. Ueki, Jay A. Nadel

Abstract

Neutrophil products are implicated in hypersecretory airway diseases. To determine the mechanisms linking a proteolytic effect of human neutrophil elastase (HNE) and mucin overproduction, we examined the effects of HNE on MUC5AC mucin production in human airway epithelial (NCI-H292) cells. Stimulation with HNE for 5–30 min induced MUC5AC production 24 h later, which was prevented by HNE serine active site inhibitors, implicating a proteolytic effect of HNE. MUC5AC induction was preceded by epidermal growth factor receptor (EGFR) tyrosine phosphorylation and was prevented by selective EGFR tyrosine kinase inhibitors, implicating EGFR activation. HNE-induced MUC5AC production was inhibited by a neutralizing transforming growth factor-α (TGF-α, an EGFR ligand) antibody and by a neutralizing EGFR antibody but not by oxygen free radical scavengers, further implicating TGF-α and ligand-dependent EGFR activation in the response. HNE decreased pro-TGF-α in NCI-H292 cells and increased TGF-α in cell culture supernatant. From these results, we conclude that HNE-induced MUC5AC mucin production occurs via its proteolytic activation of an EGFR signaling cascade involving TGF-α.

  • human airway epithelium
  • neutrophils
  • epithelial differentiation
  • MUC5AC
  • transforming growth factor-α

mucus overproduction is an important feature of chronic inflammatory airway diseases, including cystic fibrosis (5), chronic bronchitis (36), bronchiectasis (31), and severe asthma (1). Neutrophil elastase is increased in the airways of subjects with mucus overproduction, such as smokers (22), and in the airways of many individuals with inflammatory airway diseases (13, 14, 18, 44). Human neutrophil elastase (HNE) is a serine protease with potent proteolytic activity (3), which is reported to cause mucin secretion (degranulation) (7, 24, 43, 45). HNE has also been reported to cause progressive accumulation of secretory granules in hamster airway surface epithelial cells (7) and MUC5AC expression in cultured airway epithelial cells (50). However, the mechanisms linking the proteolytic effects of neutrophil elastase (presumably extracellularly) and mucin production remain unknown.

Recently, the activation of the epidermal growth factor receptor (EGFR) signaling cascade has been reported to cause mucin synthesis (46) and has been shown to be a convergent pathway through which many stimuli induce mucin expression (28, 41,46-48). However, it is not known whether an EGFR signaling cascade mediates neutrophil elastase-induced mucin production.

Transforming growth factor-α (TGF-α) is an EGFR ligand that is expressed in airway epithelial cells (19, 25). TGF-α is initially synthesized as a precursor, pro-TGF-α (11,27). The precursor is then processed by an elastase-like enzyme (8, 49), which cleaves pro-TGF-α and releases mature, soluble TGF-α (27, 32, 49). HNE also cleaves pro-TGF-α and releases mature, soluble TGF-α (34).

We hypothesized that HNE induces mucin production via activation of an EGFR cascade involving proteolytic cleavage of pro-TGF-α on the surface of airway epithelial cells, with the release of soluble TGF-α, which binds to and activates EGFR on epithelial cells, resulting in mucin production. To examine this hypothesis, we studied the effect of purified HNE on MUC5AC mucin production and the role of TGF-α in HNE-induced MUC5AC production in human airway epithelial (NCI-H292) cells.

METHODS

NCI-H292 cells, a human pulmonary mucoepidermoid carcinoma cell line, were grown in RPMI 1640 medium containing 10% FCS, penicillin (100 U/ml), streptomycin (100 μg/ml), and HEPES (25 mM) at 37°C in a humidified 5% CO2 water-jacketed incubator. After the cells reached confluence, they were serum starved for 24 h in the same medium except without FCS (serum-free medium), a condition that maintained a low basal level of MUC5AC production.

Culture condition used to examine the effect of HNE on the production of MUC5AC.

We performed ELISA measurements of MUC5AC protein in cell lysate and in cell culture supernatant. Serum-starved NCI-H292 cells were stimulated with HNE (10−7 M; Elastin Products, Owensville, MO) for 5, 10, or 30 min with TGF-α (human recombinant TGF-α, 5 ng/ml; Genzyme, Cambridge, MA) for 30 min (as positive control) or with serum-free medium alone for 30 min. After stimulation, the cells were washed three times with serum-free medium to remove the stimulus and cultured in fresh serum-free medium for another 24 h. Then ELISA measurements of MUC5AC were performed. In preliminary experiments, we found that a concentration of 10−7 M HNE had maximal effects on MUC5AC production without evidence of cell damage or cell sloughing. This concentration is lower than that observed in the sputum of patients with cystic fibrosis (13, 18).

Culture conditions used to examine the effect of HNE on the cleavage and release of TGF-α.

We studied the effect of HNE on the TGF-α concentration in the cell culture supernatant. Serum-starved NCI-H292 cells were stimulated with HNE (10−7 M) or maintained in serum-free medium alone for 30 min. To block EGFR-ligand binding, we added a neutralizing anti-human EGFR antibody (Ab-3, 1.00 μg/ml; Calbiochem, San Diego, CA) to the medium 30 min before adding HNE. Then ELISA measurements of TGF-α in the cell culture supernatant were performed. We also studied the effect of HNE on the immunocytochemical localization of TGF-α in NCI-H292 cells under two conditions. First, serum-starved NCI-H292 cells were incubated with HNE (10−7 M) or serum-free medium alone for 30 min. Then the cells were washed, and immunocytochemical staining was performed. Second, because TGF-α staining was sparse in the serum-starved condition, we also performed immunocytochemical localization of TGF-α when TGF-α was upregulated in the cells: TGF-α expression has been shown to be increased by preincubation with exogenous TGF-α (9, 10). Therefore, we incubated serum-starved NCI-H292 cells for 24 h with TGF-α (5 ng/ml) and found that the TGF-α staining in the control state was increased. Then we washed the cells three times and incubated the cells either with HNE (10−7 M) or with serum-free medium alone for 30 min, and we prepared the cells for immunocytochemical staining (“upregulated state”). To confirm HNE-induced pro-TGF-α cleavage and TGF-α release, we incubated serum-starved NCI-H292 cells with serum-free medium alone or HNE (10−7 M) for 30 min and examined changes in pro-TGF-α in cell lysate and changes in TGF-α in cell culture supernatant by immunoprecipitation followed by SDS-PAGE and silver staining. To prevent the binding of released TGF-α to EGFR, we added a neutralizing EGFR antibody (Ab-3, 1.00 μg/ml; Calbiochem) to the medium 30 min before the addition of HNE. We hypothesized that this would increase the measured TGF-α in cell culture supernatant after treatment with HNE.

Effects of elastase inhibitors, EGFR ligand antibodies, kinase inhibitors, neutralizing EGFR antibody, and oxygen free radical scavengers on HNE- or TGF-α-induced responses.

To study the effect of inhibitors of neutrophil elastase, we preincubated HNE (final concentration, 10−7 M) with ICI 200,355 (a selective neutrophil elastase inhibitor, final concentration, 10−4 M; generously provided by Zeneca Pharmaceuticals Group, Wilmington, DE) (42) or with secretory leukocyte protease inhibitor (SLPI, final concentration, 10−5 M; R & D Systems, Minneapolis, MN) (38,39) at 37°C for 30 min and then used it as a stimulus. To examine the effects of neutralizing EGFR ligand antibodies, we added anti-human TGF-α rabbit polyclonal antibody (H-50, 4.00 μg/ml; Santa Cruz Biotechnology, Santa Cruz, CA) or anti-human epidermal growth factor (EGF) rabbit polyclonal antibody (Ab-3, 4.00 μg/ml; Calbiochem) to the medium 30 min before the stimulation with HNE (10−7 M). To examine the effects of kinase inhibitors on HNE-induced MUC5AC production, we added BIBX1522 (a selective inhibitor of EGFR tyrosine kinase, 10−5 M; generously provided by Boehringer Ingelheim Pharma, Ingelheim, Germany), tyrphostin AG-1478 [a selective inhibitor of EGFR tyrosine kinase, 10−5 M; Calbiochem (30)], PD-98059 [a selective mitogen-activated protein kinase kinase (MEK) inhibitor, 10−5 M; Calbiochem (12)], or tyrphostin AG-1295 [a selective inhibitor of platelet-derived growth factor receptor (PDGFR) tyrosine kinase, 10−5 M; Calbiochem (26), negative control] to the medium 30 min before the stimulation with HNE (10−7 M). These kinase inhibitors were maintained in the medium during the 30-min stimulation period and then removed or maintained in the medium during the remaining culture period. To examine the effect of a neutralizing EGFR antibody or of oxygen free radical scavengers on HNE-induced or TGF-α-induced MUC5AC production, we added an anti-human EGFR monoclonal antibody (Ab-3, 0.25, 1.00, 4.00 μg/ml; Calbiochem) (16, 23), DMSO (1%; Sigma, St. Louis, MO), 1,3-dimethyl-2-thiourea (DMTU, 50 mM; Sigma), or superoxide dismutase (SOD, 300 U/ml; Sigma) to the medium 30 min before adding the stimulus, maintained each in the medium during the stimulation period with HNE (10−7 M) or with TGF-α (5 ng/ml), and then removed it by washing three times. As a control to determine the selectivity of the action of the neutralizing EGFR antibody, an anti-human interleukin-8 neutralizing monoclonal antibody (AHC0083, 4.00 μg/ml; Biosource International, Camarillo, CA) or anti-human tumor necrosis factor-α (TNF-α) receptor monoclonal antibody (4.00 μg/ml; Genzyme) was used instead of the neutralizing EGFR antibody.

Immunoassay of MUC5AC.

In NCI-H292 cells, newly produced mucins are released from the cell constitutively. The total production of mucins consists of both mucins in the cell lysate and mucins secreted into the cell culture supernatant. Therefore, in the present studies, the total production of MUC5AC mucin protein was calculated as the sum of the MUC5AC in cell lysate and in cell culture supernatant. Because the relative amount of MUC5AC in cell lysate and cell culture supernatant was not significantly different among various culture conditions examined in the present studies, the total amount of MUC5AC in cell lysate and cell culture supernatant was used for the analysis. For measurement of MUC5AC mucin protein, NCI-H292 cells were grown in 24-well culture plates. MUC5AC protein in cell lysate and cell culture supernatant was measured by ELISA as described previously (46). We collected cell lysates by lysing the cells with lysis buffer [1% Triton X-100, 1% deoxycholic acid, and proteinase inhibitors (Complete Mini; Roche, Indianapolis, IN)]. Fifty microliters of cell lysate or cell culture supernatant were incubated with bicarbonate-carbonate buffer (50 μl) at 40°C in a 96-well plate (Maxisorp Nunc; Fisher Scientific, Santa Clara, CA) until dry. Plates were washed with PBS and blocked with 2% BSA, fraction V (Sigma), for 1 h at room temperature. Plates were again washed and then incubated with 50 μl of MUC5AC monoclonal antibody (clone 45M1, 1:500 dilution; Neo Markers, Union City, CA) in PBS containing 0.05% Tween 20. After 1 h, the plates were washed, and 100 μl horseradish peroxidase-goat anti-mouse immunoglobulin G conjugate (1:10,000 dilution; Sigma) were dispensed into each well. After 1 h, plates were washed with PBS. Color reaction was developed with 3,3′,5,5′-tetramethylbenzidine peroxidase solution (Kirkegaard & Perry Laboratories, Gaithersburg, MD) and stopped with 2 N H2SO4. Absorbance was read at 450 nm. The amount of MUC5AC was calculated with bovine submaxillary gland mucin (type I; Sigma) as a standard. HNE causes mucin degradation (24). Therefore, allowing HNE to remain in the medium for 24 h of incubation could affect the measured mucin in the cell culture supernatant. However, we found that incubation of NCI-H292 cells with HNE for 30 min was sufficient to induce a marked increase in MUC5AC production. Thus by removing HNE by washing after a 30-min stimulation period and by measuring MUC5AC in cell lysate and cell culture supernatant 24 h later, we were able to avoid the potential degradative effect of HNE on MUC5AC.

Immunoassay of TGF-α.

For measurement of TGF-α, NCI-H292 cells were grown in six-well culture plates. TGF-α protein in cell culture supernatant was measured using the TGF-α ELISA kit (Oncogene Research Products, Boston, MA), according to the manufacturer's instructions. To determine whether HNE affects the measurement of TGF-α (e.g., by cleavage), we incubated samples of TGF-α obtained from the manufacturer with or without HNE. After 30 min of incubation with HNE (10−7 M), the measured concentration of TGF-α was unchanged. A similar lack of digestion of mature TGF-α has been reported using pancreatic elastase (21).

Immunocytochemical staining of TGF-α.

Expression of TGF-α protein was studied by immunocytochemistry: NCI-H292 cells were grown on eight-well chamber slides and incubated as described above. On completion of incubation, the cells were fixed with 4% paraformaldehyde for 30 min. Cells were then treated with 0.3% H2O2-methanol to quench endogenous peroxidase and incubated with anti-human TGF-α monoclonal antibody (Ab-2, 1:200 dilution; Calbiochem; the epitope of this TGF-α monoclonal antibody resides in the carboxyl-terminal 17-amino acid sequence of human TGF-α). After removing excess antibody by washing with PBS, we incubated cells with biotinylated horse anti-mouse immunoglobulin G (1/200 dilution; Vector Laboratories, Burlingame, CA) for 1 h at room temperature. Bound antibody was visualized according to standard procedure for avidin-biotin-peroxidase complex method (Elite, ABC kit; Vector Laboratories). Counterstaining was performed with hematoxylin.

Immunoblotting for EGFR phosphorylation.

NCI-H292 cells grown in six-well culture plates were serum starved for 24 h and then stimulated with HNE (10−7 M), TGF-α (5 ng/ml), or serum-free medium alone for 5, 10, 30, 60, or 120 min. In inhibition studies, a selective EGFR tyrosine kinase inhibitor [AG-1478, 10−5 M; Calbiochem (30)] was added to the medium 30 min before the 30-min stimulation period with HNE. Cells were then lysed on ice in PBS lysis buffer containing 1% Triton X-100, 1% deoxycholic acid, 50 mM NaF, 1 mM sodium orthovanadate, proteinase inhibitors (Complete Mini; Roche), and ICI 200,355 (10−5 M). Insoluble debris was removed by microcentrifugation (10,000 rpm) for 20 min at 4°C, and immunoprecipitation was performed. Briefly, aliquots of cell lysates containing equal amounts of protein were incubated with 4 μg of anti-EGFR antibody (Ab-5; Calbiochem) and 40 μl of protein A-agarose beads (Santa Cruz Biotechnology) for 16 h at 4°C. The beads were washed three times with PBS and resuspended and boiled in 50 μl of Laemmli sample buffer (Bio-Rad Laboratories, Hercules, CA). Then 20 μl of the sample proteins were separated by SDS-PAGE in 7.5% acrylamide gel. The resulting gel was equilibrated in the transfer buffer [25 mM Tris · HCl, 192 mM glycine, 20% (vol/vol) methanol, pH 8.3]. The proteins were then transferred electrophoretically to polyvinylidene difluoride (PVDF) membranes (Bio-Rad Laboratories), which were incubated with 5% fat-free skim milk in PBS containing 0.05% Tween 20 for 1 h and then incubated with anti-phosphotyrosine monoclonal antibody (PY99, 2 μg/ml; Santa Cruz Biotechnology) for 1 h. Bound antibody was visualized according to a standard protocol for the avidin-biotin-alkaline phosphatase complex method (ABC kit; Vector Laboratories), and densitometry was performed with the public domain NIH Image program (developed at the National Institutes of Health and available by anonymous file transfer program from zippy.nimh.gov. or floppy disk from the National Technical Information Service, Springfield, VA, part number PB95–500195GEI). As a positive control for EGFR tyrosine phosphorylation, cell lysates from unstimulated A431 cells were examined (47).

Pro-TGF-α cleavage and TGF-α release.

Serum-starved NCI-H292 cells cultured in 150-cm2 culture flasks were stimulated with serum-free medium alone or HNE (10−7 M) for 30 min. To upregulate baseline expression of TGF-α, we incubated cells with TGF-α (5 ng/ml) for 30 min before the serum starvation period. To facilitate the detection of TGF-α released into cell culture supernatant by preventing EGFR-ligand binding, we added a neutralizing EGFR antibody (Ab-3, 1.00 μg/ml; Calbiochem) to the culture medium 30 min before adding HNE. When incubation was completed, the cell culture supernatants were collected and desalted by dialysis against distilled water with cellulose ester membrane (Spectra/Por CE Membrane MWCO: 500; Spectrum, Rancho Dominguez, CA), and then the supernatants were concentrated by lyophilization. The cell lysates were collected on ice with PBS lysis buffer containing 1% Triton X-100, 1% deoxycholic acid, 50 mM NaF, 1 mM sodium orthovanadate, and proteinase inhibitors (Complete Mini; Roche). Aliquots of samples (the cell lysates or the concentrated cell culture supernatants) containing equal amounts of protein were incubated with 2 μg of rabbit anti-human TGF-α antibody (H-50; Santa Cruz Biotechnology) and 30 μl of Protein A/G PLUS-Agarose (Santa Cruz Biotechnology) for 16 h at 4°C (immunoprecipitation). The beads were washed seven times with PBS and resuspended and boiled in Laemmli sample buffer (Bio-Rad Laboratories). Immunoprecipitated proteins were then separated by SDS-PAGE in 12% acrylamide gel, and the resultant gel was stained by a standard silver staining method.

Counting of cell number after incubation.

To examine whether effects of HNE were due to changes in cell number, we counted the numbers of NCI-H292 cells incubated in 24-well culture plates. After completion of the incubation under various conditions (described above), the cells were washed with calcium-magnesium-free PBS, disaggregated with trypsin-EDTA solution, and resuspended in PBS. The number of live cells was counted by the trypan blue dye exclusion method, using a hemocytometer. Cell numbers after incubation under the various conditions studied were not significantly different.

Statistical analysis.

For analysis of the results of ELISA measurements of MUC5AC protein, one-way ANOVA for repeated measurements was used on raw data to determine statistically significant differences among groups. When statistical significance was identified in the ANOVA, a Student-Newman-Keuls test was used for multiple comparisons. Data obtained from ELISA measurements of TGF-α protein were not normally distributed and were analyzed by the nonparametric Wilcoxon test. A probability <0.05 for the null hypothesis was accepted as indicating a statistically significant difference. Because amounts of MUC5AC produced by NCI-H292 cells were variable depending on the passage number of the cells used, the percentage above control value was used to report MUC5AC data. Measurements of TGF-α protein are expressed as picograms per milliliter. All data are expressed as means ± SE.

RESULTS

HNE increases MUC5AC mucin production via a proteolytic action.

When NCI-H292 cells were incubated with HNE (10−7 M) for variable periods (5, 10, or 30 min) and then washed to remove the HNE, MUC5AC protein production over the next 24 h was increased markedly (Fig. 1 A); MUC5AC production induced by exposure to HNE for as brief a time as 5 min was not significantly different from the induced MUC5AC production by a 30-min exposure to HNE (Fig. 1 A). HNE-induced mucin production was prevented by preincubation of HNE with inhibitors of HNE proteolytic activity (10−4 M ICI 200,355 or 10−5 M SLPI, Fig. 1 B), indicating that HNE-induced mucin production was due to an effect of the catalytic site of the enzyme. ICI 200,355 alone (10−4 M) or SLPI alone (10−5 M) had no significant effect on MUC5AC production (P > 0.05, n = 5 separate experiments; ICI 200,355 alone, +18.8 ± 6.9% above control; SLPI alone, +4.9 ± 13.7% above control).

Fig. 1.

ELISA for MUC5AC protein in NCI-H292 cells. A: effect of neutrophil elastase on MUC5AC production. Serum-starved NCI-H292 cells were stimulated with serum-free medium alone for 30 min (control) or with human neutrophil elastase (HNE, 10−7 M) for 5, 10, or 30 min, washed three times to remove the stimulus, and cultured in fresh serum-free medium for another 24 h. Then the cell lysate and cell culture supernatant were collected, and MUC5AC was measured by ELISA. Data are expressed as % above control; means ± SE; n = 5 separate experiments; *P< 0.001, significantly different from control. B: effect of neutrophil elastase inhibitors on neutrophil elastase-induced MUC5AC production. Serum-starved NCI-H292 cells were stimulated with serum-free medium alone (control), HNE (10−7 M) alone, ICI 200,355 (ICI, 10−4 M) plus HNE (10−7 M), or secretory leukocyte inhibitor (SLPI, 10−5 M) plus HNE (10−7 M) for 30 min, washed three times to remove the stimulus, and cultured for another 24 h in fresh serum-free medium. (For details, see methods.) Then the cell lysate and cell culture supernatant were collected, and MUC5AC was measured by ELISA. Data are expressed as % above control; means ± SE;n = 5 separate experiments; *P < 0.001, significantly different from control; †P < 0.001, significantly different from HNE alone.

EGFR and MEK activation mediate HNE-induced MUC5AC production.

Addition of selective EGFR tyrosine kinase inhibitors (10−5 M BIBX1522 or 10−5 M tyrphostin AG-1478) to NCI-H292 cells 30 min before the addition of HNE and maintained in the medium during the remaining culture period prevented HNE-induced MUC5AC production (Fig.2 A). These inhibitory effects by selective EGFR tyrosine kinase inhibitors were observed even if the inhibitors were removed after stimulation period with HNE (Fig.2 B). Similarly, a selective MEK inhibitor (PD-98059, 10−5 M) prevented the response (Fig. 2, A andB). However, the PDGFR tyrosine kinase inhibitor (tyrphostin AG-1295, 10−5 M; negative control) showed no inhibitory effect (Fig. 2, A and B). These results implicate activation of an EGFR signaling cascade in HNE-induced MUC5AC production.

Fig. 2.

Effect of tyrosine kinase inhibitors on MUC5AC production induced by neutrophil elastase. Serum-starved NCI-H292 cells were stimulated with serum-free medium alone (control) or with HNE (10−7 M) for 30 min, washed three times to remove the stimulus, and cultured in fresh serum-free medium for another 24 h. BIBX1522 (10−5 M), tyrphostin AG-1478 (10−5 M), PD-98959 (10−5 M), or tyrphostin AG-1295 (10−5 M) was added to the medium 30 min before addition of HNE and maintained in the medium during the remaining culture period (A) or only for the 30-min stimulation period with HNE (B). Then the cell lysate and cell culture supernatant were collected, and MUC5AC was measured by ELISA. Data are expressed as % above control; means ± SE; n = 5 separate experiments; *P < 0.001, significantly different from control; †P < 0.001, significantly different from HNE; §P > 0.05, not significantly different from HNE alone.

To confirm the role of EGFR tyrosine kinase phosphorylation in the response to HNE, we incubated serum-starved NCI-H292 cells with HNE (10−7 M) for 5, 10, 30, 60, and 120 min, and we performed immunoblotting analysis for phosphorylated EGFR: HNE increased phosphorylated EGFR, an effect that was most prominent after 30 min of incubation with HNE (Fig. 3 A), and this induction after 30 min of incubation with HNE was prevented by a selective inhibitor of EGFR tyrosine kinase (AG-1478, 10−5 M; Fig. 3 B). Relative intensity of phosphorylated EGFR measured by densitometry was as follows: TGF-α: +335.0 ± 25.2% above control; HNE (30 min): +173.9 ± 20.7% above control; AG-1478 plus HNE (30 min): −3.2 ± 14.4% above control (n = 3 separate experiments).

Fig. 3.

Immunoblotting for phosphorylated epidermal growth factor receptor (EGFR). A: serum-starved NCI-H292 cells were stimulated with serum-free medium alone for 30 min (control) or with HNE (10−7 M) for 5, 10, 30, 60, or 120 min and lysed. Immunoprecipitation for EGFR was performed with cell lysates containing equal amounts of protein. The sample proteins were then separated by SDS-PAGE in 7.5% acrylamide gel, transferred electrophoretically to polyvinylidene difluoride (PVDF) membranes, and blotted with anti-phosphotyrosine monoclonal antibody, as described inmethods. Results are representative of 3 separate experiments. B: serum-starved NCI-H292 cells were stimulated with serum-free medium alone (control), transforming growth factor-α (TGF-α, 5 ng/ml), HNE alone (10−7 M), or a selective EGFR tyrosine kinase inhibitor [AG-1478 (10−5 M)] plus HNE (10−7 M) for 30 min and lysed. Immunoprecipitation for EGFR was performed with cell lysates containing equal amounts of protein. The sample proteins were then separated by SDS-PAGE in 7.5% acrylamide gel, transferred electrophoretically to PVDF membranes, and blotted with anti-phosphotyrosine monoclonal antibody, as described inmethods. A431 cells were used as positive control for phosphorylated EGFR. Results are representative of 3 separate experiments.

Neutralizing EGFR antibody prevents HNE-induced MUC5AC production.

Addition of a neutralizing EGFR antibody (0.25, 1.00, 4.00 μg/ml) 30 min before adding HNE and maintaining the antibody in the medium during the 30-min stimulation period with HNE prevented HNE-induced MUC5AC production dose-dependently and completely (at a concentration of 4.00 μg/ml; Fig. 4). However, neither a neutralizing interleukin-8 antibody (4.00 μg/ml; Fig. 4) nor a neutralizing TNF-α receptor antibody (4.00 μg/ml; data not shown) had an inhibitory effect on HNE-induced MUC5AC production.

Fig. 4.

Effect of neutralizing EGFR antibody, TGF-α antibody, or scavengers of oxygen free radicals on MUC5AC production induced by neutrophil elastase or by TGF-α. Serum-starved NCI-H292 cells were stimulated with serum-free medium alone (control), TGF-α (5 ng/ml), or HNE (10−7 M) for 30 min, washed 3 times to remove the stimulus, and cultured in fresh serum-free medium for another 24 h. A neutralizing EGFR antibody [0.25 μg/ml, EGFR Ab (0.25); 1.00 μg/ml, EGFR Ab (1); 4.00 μg/ml, EGFR Ab (4)], a TGF-α antibody (4.00 μg/ml, TGF-α Ab), DMSO (1%), 1,3-dimethyl-2-thiourea (DMTU, 50 mM), or superoxide dismutase (SOD, 300 U/ml) was added to the medium 30 min before addition of HNE. (For details, see methods.) Then the cell lysate and cell culture supernatant were collected, and MUC5AC was measured by ELISA. IL-8, interleukin-8. Data are expressed as % above control; means ± SE; n = 5 separate experiments; *P< 0.001, significantly different from control; †P < 0.001, significantly different from TGF-α alone; ‡P< 0.001, significantly different from HNE alone; §P> 0.05, not significantly different from HNE alone.

Oxygen free radical scavengers do not inhibit HNE-induced MUC5AC production.

Because oxidative stress is reported to induce MUC5AC production in NCI-H292 cells via transactivation of EGFR (47) and because DMSO (1%) is reported to inhibit hydrogen peroxide-induced but not TGF-α-induced MUC5AC production (47), we also examined the effects of scavengers of oxygen free radicals: addition of oxygen free radical scavengers (1% DMSO, 50 mM DMTU, or 300 U/ml SOD) 30 min before adding HNE and maintaining the scavenger in the medium during the 30-min stimulation period with HNE did not inhibit HNE-induced MUC5AC production (Fig. 4).

TGF-α plays a role in HNE-induced MUC5AC production.

A neutralizing EGFR antibody, but not oxygen free radical scavengers, blocked HNE-induced MUC5AC production, suggesting that ligand-dependent EGFR activation is involved in the response. TGF-α is present in airway epithelial cells (19, 25), and TGF-α induces MUC5AC production (46), which is confirmed in the present studies (Fig. 4). These findings suggest that TGF-α could play a role in HNE-induced mucin production. Addition of a neutralizing TGF-α antibody (4.00 μg/ml) 30 min before adding a stimulus (HNE or TGF-α) inhibited both HNE (Fig. 4)- and TGF-α (data not shown)-induced MUC5AC production, implicating TGF-α in the response to HNE. Addition of a neutralizing EGF antibody (4.00 μg/ml) had no effect on HNE-induced MUC5AC production (n = 5 separate experiments; HNE alone: +271.9 ± 16.3% above control,P < 0.001 significantly different from control; EGF antibody plus HNE: +277.5 ± 57.1% above control,P > 0.05 not significantly different from HNE alone).

HNE induces cleavage of pro-TGF-α and release of TGF-α from NCI-H292 cells.

The proteolytic effect of HNE on TGF-α processing was examined in NCI-H292 cells in three ways. First, we measured the concentration of TGF-α in the cell culture supernatant by ELISA: incubation of serum-starved NCI-H292 cells with HNE (10−7 M) for 30 min increased the TGF-α protein concentration in the cell culture supernatant from 2.70 ± 1.03 pg/ml to 4.52 ± 1.44 pg/ml (n = 6 separate experiments; P < 0.05), suggesting that TGF-α was released from the cells. The HNE-induced increase in TGF-α concentration in the cell culture supernatant was more prominent when EGFR-ligand binding was blocked by a neutralizing EGFR antibody (n = 6 separate experiments, EGFR antibody plus HNE: 17.36 ± 5.94 pg/ml,P < 0.05 significantly different from HNE alone: 5.12 ± 1.94 pg/ml). Second, we examined the effect of HNE on TGF-α immunostaining: we found weak immunostaining of TGF-α in serum-starved NCI-H292 cells, which was decreased by incubation with HNE (10−7 M) for 30 min (Fig.5 A). In additional studies, baseline TGF-α staining was increased by preincubation with TGF-α (5 ng/ml, 24 h; Fig. 5 B). After upregulation of TGF-α staining, subsequent addition of HNE (10−7 M) to NCI-H292 cells for 30 min induced a marked decrease in TGF-α staining (Fig.5 B), an effect that was prevented by preincubation of HNE with a selective HNE inhibitor (ICI 200,355, 10−4 M; Fig.5 B), confirming that HNE caused TGF-α release from the cells via its proteolytic effect. Third, to confirm that HNE cleaves membrane-bound TGF-α precursor (pro-TGF-α) and releases TGF-α into the culture medium, we examined changes in pro-TGF-α in cell lysate and in released TGF-α in cell culture supernatant by immunoprecipitation followed by SDS-PAGE and silver staining: incubation of NCI-H292 cells with HNE for 30 min decreased the amount of pro-TGF-α (∼17 kDa) in cell lysate (Fig.6 A) and increased the amount of TGF-α (∼5.5 kDa) in cell culture supernatant (Fig.6 B), as reported previously (34). The increase in the ∼5.5-kDa TGF-α molecule in cell culture supernatant was more prominent when EGFR-ligand binding was blocked by neutralizing EGFR antibody (Fig. 6 B). [The ∼22-kDa molecule in cell lysate (Fig. 6 A) may correspond to intermediate molecules, i.e., glycosylated pro-TGF-α species (35).] These results indicate that HNE cleaves cell surface pro-TGF-α and releases TGF-α, and they implicate the released TGF-α in ligand-dependent EGFR activation induced by HNE.

Fig. 5.

Immunocytochemical staining of TGF-α protein in NCI-H292 cells. The cells were either serum starved for 24 h (A) or preincubated with TGF-α (5 ng/ml) in serum-free medium for 24 h to increase basal level of TGF-α staining (upregulated state,B). Then the cells were washed 3 times and stimulated with serum-free medium alone (control), HNE (10−7 M), or ICI 200,355 (ICI, 10−4 M) plus HNE (10−7 M) for 30 min, fixed with 4% paraformaldehyde, and stained with TGF-α antibody. For details, see methods. Photomicrographs are representative of 3 separate experiments. Arrowheads, localization of TGF-α; bar, 20 μm.

Fig. 6.

Kinetics of pro-TGF-α cleavage and TGF-α release by neutrophil elastase. A: TGF-α precursor in cell lysate. NCI-H292 cells were deprived of serum for 24 h (serum starved). To increase baseline TGF-α expression, cells were incubated with TGF-α (5 ng/ml) for 30 min before the serum-starvation period (upregulated). The cells were then stimulated with serum-free medium alone (control) or HNE (10−7 M) for 30 min, and the cell lysates were collected. Immunoprecipitation was performed using an anti-TGF-α antibody and equal amounts of the cell lysate. The sample proteins were separated by SDS-PAGE in 12% acrylamide gel, and the resultant gel was stained by a standard silver staining method. Results are representative of 3 separate experiments. B: TGF-α released into cell culture medium. Serum-starved NCI-H292 cells were stimulated with serum-free medium alone (control) or HNE (10−7 M) for 30 min, and the cell lysates were collected. To prevent EGFR-ligand binding, we added a neutralizing EGFR antibody (1.00 μg/ml) to the culture medium 30 min before the addition of HNE (EGFR Ab + HNE). The cell culture supernatants were then collected and concentrated as described in methods. Immunoprecipitation was performed using anti-TGF-α antibody and the concentrated cell culture supernatant. The sample proteins were then separated by SDS-PAGE in 12% acrylamide gel, and the resultant gel was stained by a standard silver staining method. Recombinant TGF-α was used as a positive control (TGF-α). Results are representative of 3 separate experiments.

DISCUSSION

Neutrophils have been implicated in mucus hypersecretory diseases, and the neutrophil serine protease elastase has been shown to induce mucin production (7, 50). However, mechanisms linking the proteolytic effect of HNE and mucin overproduction are unknown.

Here we show that HNE induces marked MUC5AC production in human airway epithelial cells and that this induction is prevented by inhibitors of the serine active site of HNE [ICI 200,355 (42) and SLPI (38, 39)], implicating a proteolytic action of HNE in the response.

We hypothesized that HNE-induced mucin production is mediated via an EGFR signaling pathway. Incubation of the epithelial cells with HNE resulted in EGFR tyrosine phosphorylation and in subsequent MUC5AC production. EGFR tyrosine phosphorylation occurred as early as 5 min and was maximal at 30 min. Both the induced EGFR phosphorylation and the induced MUC5AC production were blocked when selective EGFR tyrosine kinase inhibitors were added 30 min before addition of HNE and maintained during the stimulation period with HNE, implicating rapid EGFR activation in the responses to HNE. The MEK-mitogen-activated protein kinase (MAPK) transduction pathway is known to be downstream of EGFR activation, and a selective inhibitor of MEK (PD-98059) is reported to inhibit MUC5AC production in NCI-H292 cells (47). Here we show that HNE-induced MUC5AC production is also prevented by preincubation with the MEK inhibitor (PD-98059), suggesting the MEK-MAPK pathway as downstream of EGFR phosphorylation in causing HNE-induced MUC5AC production.

Next, we investigated the mechanism of EGFR activation by HNE. EGFR can be activated by two separate processes: the binding of an EGFR ligand (e.g., TGF-α, EGF) to EGFR activates the receptor tyrosine kinase and induces tyrosine phosphorylation (20, 27, 40). Alternatively, EGFR tyrosine phosphorylation can be induced by a ligand-independent mechanism [e.g., stimulation by oxygen free radicals (17, 47)]. In the present studies, HNE-induced MUC5AC production was prevented by preincubation with a neutralizing EGFR antibody, which is known to bind to the extracellular domain of the EGFR and to inhibit ligand binding to EGFR (16, 23). On the other hand, preincubation with scavengers of oxygen free radicals (DMSO, DMTU, or SOD) failed to inhibit HNE-induced MUC5AC production. These results indicate that ligand-dependent EGFR activation plays a major role during the 30-min stimulation period with HNE. This led us to seek a proteolytic action of HNE to explain this ligand-dependent activation of EGFR.

The fact that incubation of epithelial cells with HNE for as short a time as 5 min (see Fig. 1) was sufficient to cause marked MUC5AC production suggests that an early cleavage event initiates the induction of MUC5AC production. Furthermore, because of the relatively high molecular mass of HNE [∼30 kDa (3)], it is likely that HNE causes MUC5AC production via its proteolytic action on molecules that reside on the outer surface of the epithelial cells.

The EGFR ligand TGF-α is known to be present on the surface of normal human airway epithelial cells (25) and is upregulated in inflammatory airway diseases (19). Furthermore, TGF-α is reported to induce MUC5AC production in NCI-H292 cells in vitro and in experimental animals in vivo (46). Therefore, we hypothesized that TGF-α located on the epithelial surface could be cleaved by HNE and thus act as an EGFR ligand in MUC5AC production.

TGF-α is initially synthesized as a 160 [human (11)]-amino acid membrane-bound precursor, pro-TGF-α (27). Newly synthesized pro-TGF-α is cleaved at alanine-valine sequences (11) at the amino and carboxyl terminus of mature TGF-α by elastase-like enzymes (8,49), resulting in the release of mature, soluble TGF-α (27, 49), which is more potent as an EGFR ligand than the membrane-bound form of TGF-α (6). The released TGF-α is then free to bind to EGFR on the same or nearby cells (35). The elastase-like enzymes that cleave pro-TGF-α may be produced by the TGF-α-producing cells (8, 49). In addition, in cultured fibroblasts, neutrophil elastase has been reported to cleave the TGF-α precursor that is located on the cell surface, resulting in the release of mature TGF-α, decreasing the cell surface TGF-α precursor and increasing the amount of the mature TGF-α species in the culture medium (34). Therefore, we hypothesized that HNE causes mucin production by cleaving pro-TGF-α on the outer surface of airway epithelial cells, which allows the released soluble TGF-α to activate EGFR, resulting in mucin production. We reasoned that cleavage of pro-TGF-α on the outer surface of the NCI-H292 cells by HNE would lead to an increase in the concentration of TGF-α (measured by ELISA) in the cell culture supernatant and to a decrease in the amount of pro-TGF-α in the cell lysate. However, EGFRs are expressed in NCI-H292 cells (37), and, after binding to EGFR, TGF-α is internalized rapidly and degraded (29). This results in a lower concentration of TGF-α in the cell culture supernatant. Despite this limitation, we were able to show that incubation of NCI-H292 cells with HNE for 30 min increased the TGF-α concentration in the cell culture supernatant. In addition, the elastase-induced increase in TGF-α concentration was magnified by blocking EGFR-ligand binding. These results implicate the cleavage and release of TGF-α from the surface of the NCI-H292 cells and the binding of the released TGF-α to EGFR on the surface of the cells. We also reasoned that cleavage of pro-TGF-α from the outer surface of the NCI-H292 cells by HNE should decrease pro-TGF-α on the epithelial cell surface. We showed that TGF-α immunostaining was decreased after incubation with HNE. Because TGF-α immunostaining was not strong in serum-starved NCI-H292 cells, we also prestimulated the cells with TGF-α, which is known to upregulate TGF-α expression (9, 10). This stimulation increased the TGF-α staining in the control state and exaggerated the HNE-induced decrease in TGF-α staining. Preincubation of HNE with a selective HNE inhibitor (ICI 200,355) prevented the HNE-induced decrease in TGF-α staining, further implicating a proteolytic effect. The combination of the HNE-induced decrease in TGF-α immunostaining in NCI-H292 cells and the HNE-induced increase in TGF-α concentration (ELISA) in the cell culture supernatant further implicates HNE in the release of TGF-α from the surface of NCI-H292 cells. Furthermore, by immunoprecipitation followed by SDS-PAGE, we confirmed that HNE decreased the amount of a 17-kDa pro-TGF-α molecule in cell lysate (on the surface of NCI-H292 cells) and that HNE increased the amount of a ∼5.5-kDa TGF-α molecule in cell culture supernatant. Together with the finding that a neutralizing TGF-α antibody inhibited HNE-induced MUC5AC production, the HNE-induced release of TGF-α from the epithelial cells strongly suggests that HNE-induced release of TGF-α is involved in HNE-induced EGFR activation and subsequent MUC5AC production.

Here we show that as short a period of exposure to HNE as 5 min causes marked subsequent upregulation of MUC5AC production. Feedback mechanisms could later modify the effect of the initial cleavage event induced by HNE. For example, EGFR ligands (e.g., TGF-α) are reported to upregulate TGF-α expression (4, 9, 10, 33), and an EGFR ligand (EGF) is reported to upregulate EGFR expression (4). In addition, TGF-α is reported to accelerate the cleavage of pro-TGF-α by activating EGFR (2). Such observations suggest that positive feedback mechanisms may further affect mucin production induced by short exposures to HNE.

Although EGF is an EGFR ligand and TNF-α induces EGFR expression (46) and is synthesized as a membrane-bound precursor similar to TGF-α, neither a neutralizing antibody to EGF (seeresults) nor a neutralizing antibody to TNF-α receptor (data not shown) showed an inhibitory effect on HNE-induced MUC5AC production, and TNF-α alone does not induce MUC5AC expression (46). However, other EGFR ligands not studied here could also contribute to HNE-induced mucin production.

A previous study reported that HNE-induced mucin expression in airway epithelial cells is due to a proteolytic effect of HNE (50). Mucin expression was associated with the induction of oxidative stress (15), but no proteolytic mechanism was described. In the present study, scavengers of oxygen free radicals were ineffective in preventing HNE-induced mucin production. Nevertheless, we acknowledge that many stimuli (e.g., HNE) may induce oxygen free radical production in epithelial cells. However, we found no evidence that oxygen free radicals are involved in the proteolytic action of HNE, leading to mucin production. To the contrary, the present studies show that HNE cleaves EGFR proligand, activating EGFR and inducing mucin production via an EGFR cascade.

Previous studies showed that chemoattractants cause neutrophil-dependent degranulation of goblet cells involving adhesive interactions between epithelial cells and neutrophils, with elastase activity occurring at the cell interface, causing mucus secretion (45). Similar adhesive interactions of neutrophils and epithelial cells may also play a role in HNE-induced mucin production in vivo.

Neutrophil recruitment normally provides protective effects on the host against inhaled materials. We suggest that the action of HNE on the secretion and production of mucin assists in the entrapment and clearance of potentially invasive materials. However, in pathological states, overproduction of mucins results in mucus plugging, leading to airway obstruction. Neutrophil recruitment, elastase release, and mucus hypersecretion occur in many chronic inflammatory diseases of the airways. The present studies show for the first time that a proteolytic action of neutrophil elastase induces MUC5AC mucin production via an EGFR signaling pathway. The results suggest that cleavage of the EGFR ligand TGF-α and its subsequent activation of EGFR are involved in the response. They demonstrate a novel mechanism linking neutrophil-induced proteolysis and mucin production, and they suggest novel strategies for therapeutic intervention.

Acknowledgments

We thank Drs. Pierre-Regis Burgel and Yeon-Mok Oh for useful discussions and Dominic C. Tam for technical assistance.

Footnotes

  • We also thank Boehringer Ingelheim Pharma for providing BIBX1522 and Zeneca Pharmaceuticals group for providing ICI 200,355. This work was supported by private funds.

  • Address for reprint requests and other correspondence: J. A. Nadel, Cardiovascular Research Inst., Box 0130, Univ. of California San Francisco, San Francisco, CA 94143-0130 (E-mail:janadel{at}itsa.ucsf.edu).

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

  • April 12, 2002;10.1152/ajplung.00455.2001

REFERENCES

View Abstract