Bradykinin Stimulates IL-8 Production in Cultured Human Airway Smooth Muscle Cells: Role of Cyclooxygenase Products¹

Bronchial asthma is a chronic airway inflammatory disease that is characterized by the infiltration of a large number of inflammatory cells into the airway (1). IL-8, the C-X-C chemokine, is a potent chemoattractant for neutrophils (2) as well as eosinophils (3) and has been implicated in a number of inflammatory airway diseases, such as cystic fibrosis (4), adult respiratory distress syndrome (5), and chronic bronchitis (6). Increased levels of IL-8 have been reported in the blood and bronchial mucosa (7), in the bronchoalveolar lavage fluid (8, 9), in macrophages in the bronchoalveolar lavage fluid (10), and in the bronchial epithelial cells (11) of patients with asthma. In vitro studies have identified a number of potential sources of IL-8 in the airway. These include airway epithelial cells (11), macrophages (10), mast cells (12), and fibroblasts (13). Airway smooth muscle (ASM)³ was traditionally considered to have only contractile and proliferative functions and has attracted little attention with regard to its ability to express and release inflammatory mediators. However, recent research has suggested that ASM cells could be a rich source of inflammatory mediators and could play a critical role in perpetuating the inflammatory process in the airway (14). For instance, cytokine-stimulated human ASM cells have been shown to release various cytokines such as granulocyte-macrophage CSF (GM-CSF) (15), IL-11, IL-6 (16), and RANTES (17). Human ASM cells also respond to cytokines and other stimuli by expressing the inducible form of the inflammatory gene cyclooxygenase (COX)-2 and releasing high concentrations of prostanoids (18, 19, 20).

Prostanoid synthesis is mediated by two isoforms of COX, the constitutive COX-1 and the inducible COX-2, whose expression is regulated by cytokines (21). Reports from synovial fibroblasts showing that derived CSFs are modulated by endogenous and exogenous prostanoids (22, 23) suggest that prostanoids, in addition to their role as inflammatory mediators, may be important modulators of cytokine production. Although increased prostanoid release following COX-2 induction in cultured human ASM cells has recently been reported (18, 19, 20), the effect of prostanoids on cytokine release in human ASM cells has not been investigated.

Bradykinin (BK), which is a potent inflammatory mediator in patients with bronchial asthma (24, 25, 26), has recently been shown to stimulate cytokine expression. BK increases the expression of IL-1β, IL-2, and IL-6 from isolated guinea pig lung strips (27), IL-1β from cultured human fibroblasts (28), and IL-8 from human decidua-derived cells (29). Recently, we have reported that BK, like IL-1β, causes the induction of COX-2 and the release of prostanoids from human ASM cells (19). Since IL-1β reportedly causes IL-8 release from human ASM cells (16) and PGE₂ enhances IL-8 and IL-6 production by IL-1-stimulated human synovial fibroblasts (23), we postulated that BK might also cause IL-8 release from human ASM cells, and that the high concentration of prostanoids following COX-2 induction could play a modulating role in this process.

The present study investigated whether BK could cause IL-8 production in human ASM cells and determined whether COX products and COX-2 isoenzyme induction were involved in this process.

Human tracheas were obtained from two postmortem individuals (one 44-yr-old male and one 52-yr-old female) within 12 h of death; neither individual had any evidence of airway diseases. Primary cultures of human ASM cells were prepared from explants of ASM according to previously reported methods (18, 19). Cells at passage three to four were used for all experiments. We have shown previously that cells grown in this manner depict the immunohistochemical and light microscopic characteristics of typical ASM cells (18).

The cells were cultured to confluence in 10% FCS (Seralab, Crowly Down, U.K.)/DMEM (Sigma, Poole, U.K.) in humidified 5% CO₂/95% air at 37°C in 24-well culture plates and growth-arrested in serum-deprived medium for 24 h before experiments. Immediately before each experiment, fresh serum-free medium containing BK (Sigma) was added. In the time-course experiments, the cells were incubated with BK (10 μM) for 1 to 24 h, whereas in the concentration response experiments, the cells were incubated for 16 h with 0.01 to 100 μM of BK. In most experiments thereafter, the cells were incubated with 10 μM of BK for 16 h. At the indicated times, the culture media were harvested and stored at −20°C until the RIA for PGE₂ content as a representative of prostanoid generation (18) and/or an ELISA for IL-8. The anti-PGE₂ antiserum (Sigma) had negligible cross-reactivity in our study (18). To test the inhibition of various drugs on the effect of BK, the nonselective COX inhibitors indomethacin (IND) and flurbiprofen (FLU), the protein synthesis inhibitors cycloheximide (CHX) and actinomycin D (ACT), the antiinflammatory steroid dexamethasone (DEX), the B₁ receptor antagonist desArg⁹, (Leu⁸)-BK (Sigma), the selective COX-2 inhibitor NS-398 (N-(2-cyclohexyloxy-4-nitrophenyl)-methanesulfonamide) (Cayman Chemical, Ann Arbor, MI), and the B₂ receptor antagonist DArg(Hyp³,Thi⁵,Dtic⁷,Oic⁸)-BK (HOE-140) (a kind gift of Drs. R. N. Zahlten and B. A. Scholkens, Hoechst Aktiengesellschaft, Frankfurt, Germany) were added 30 min before the addition of BK. Experiments with the selective BK B₁ receptor agonist desArg⁹-BK, the B₂ receptor agonist (Tyr(Me)⁸)-BK, COX substrate arachidonic acid (AA), calcium ionophore A23187, and exogenous PGE₂ (all from Sigma) were conducted in the same way as those for BK.

The concentration of IL-8 in the culture medium was determined by ELISA (CLB, Amsterdam, The Netherlands) according to the manufacturer’s instructions. Briefly, ELISA plates were coated overnight at room temperature with 200 μL of anti-human IL-8-coating Ab that had been diluted in 0.1 M carbonate/bicarbonate buffer (pH 9.6). Plates were then washed five times with PBS (pH 7.2–7.4) containing 0.05% Tween 20 and blocked for 1 h at room temperature with 200 μL of blocking buffer. Plates were washed again, and 100 μL of samples containing standard amounts of human rIL-8 as well as study samples were added in duplicate to individual wells and incubated at room temperature for 1 h. After five washes, 100 μL of biotinylated IL-8 Ab diluted in dilution buffer was added for 1 h. After another five washes, 100 μL of streptavidin-horseradish peroxidase (HRP) conjugate that had been diluted to 1/10,000 in dilution buffer was added for 30 min. After a final wash, 100 μL of the substrate buffer containing the HRP substrate tetramethylbenzidine dihydrochloride and hydrogen peroxide in 0.05 M phosphate-citrate buffer (pH 5.0) was added for 30 min in the dark and color-developed in proportion to the amount of IL-8 present. The reaction was stopped by adding 100 μL of stop solution (1.8 M sulfuric acid), and the degree of color that had been generated was determined by measuring the OD at 450 nm in a Dynatech MR5000 microplate reader (Billinghurst, U.K.). The standard curve was linearized and subjected to regression analysis. The IL-8 concentration of unknown samples was extracted using the standard curve. The results were expressed as picograms per milliliter of culture medium. The sensitivity of the ELISA kit in our study was at least 5 pg/ml, which was consistent with the manufacturer’s specifications. According to the kit insert, the anti-IL-8 Ab does not cross-react with IL-1 through IL-7, IL-9 through IL-11, TNF, IFN-γ, GM-CSF, and RANTES. All of the reagents used in the assay were supplied by the ELISA manufacturer, with the exception of the HRP substrate tetramethylbenzidine dihydrochloride, which was obtained from Sigma.

The toxicity of all the chemicals used in this study as well as their vehicles (dimethyl sulfoxide and ethanol) (Sigma, final concentration 1.0% v/v) to human ASM cells was determined by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay (18, 19). After a 24-h incubation with the chemicals, 20 μl of 5 mg/ml thiazolyl blue (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) (Sigma) was added to the culture medium in 96-well plates and incubated for 1 h at 37°C. After removing the medium, 200 μl of DMSO was added to solubilize the blue-colored tetrazolium, the plates were shaken for 5 min, and the OD₅₅₀ values were read in a Dynatech MR5000 microplate reader. Viability was set as 100% in control cells.

Data were expressed as the mean ± SEM from n determinations. The statistical analysis was performed using the statistical software from SPSS Inc. (Chicago, IL) (30). A one-way ANOVA and/or an unpaired, two-tailed Student’s t test were used to determine the significant differences between the means. The results were adjusted for multiple testing using Bonferroni’s correction. p values of <0.05 were accepted as statistically significant.

To investigate the time course of IL-8 production, human ASM cells were cultured in the presence or absence of BK (10 μM). Cell culture supernatants from control cells were collected at 2, 8, and 24 h; supernatants from BK-treated cells were collected at 0.25, 0.5, 1, 2, 4, 8, 16, and 24 h. As shown in Figure 1,A, basal (0 h) IL-8 release was 5.92 pg/ml. There was a slight increase in IL-8 production from control cells over the 24-h incubation (6.15 pg/ml at 2 h, 6.48 pg/ml at 8 h, and 11.34 pg/ml at 24 h). There was a marked and time-dependent increase in IL-8 release following stimulation with BK; a significant difference was observed after 2 h of stimulation as compared with IL-8 production from control cells (p < 0.01), and the highest IL-8 concentration was achieved after 16 h of stimulation (p < 0.001). When the cells were cultured with BK at concentrations of 0.01, 0.1, 1.0, 10, and 100 μM for 16 h, a concentration-dependent increase in IL-8 production was also observed that was significant from 0.01 μM (p < 0.001) (Fig. 1 B).

The effect of the nonselective COX inhibitor IND, the selective COX-2 inhibitor NS-398, the protein synthesis inhibitors CHX and ACT, and the steroid DEX was assessed on BK-induced PGE₂ release and IL-8 production from human ASM cells. As shown in Figure 2,A, BK (10 μM) caused a marked increase in PGE₂ release; this increase was inhibited in a strong and concentration-dependent manner by both IND and NS-398 and was abolished at a concentration of 10 μM. Interestingly, BK-induced IL-8 production was also markedly inhibited by the same concentration range of IND and NS-398 in a concentration-dependent fashion, although production could not be abolished by either inhibitor (Fig. 2,B); this finding suggests that BK-induced IL-8 production is mediated by the COX products, including those from the inducible COX-2 isoform, after BK treatment. We have shown before that BK-induced long-term (4 h) PGE₂ release and COX-2 induction are inhibited by CHX, ACT, and DEX (19); BK-induced IL-8 production was also abolished by CHX, ACT, and DEX in this past study (all 1.0 μM, Fig. 3).

If COX products are involved in BK-stimulated IL-8 production, then such products should be able to cause IL-8 production on their own. To prove this, we examined whether the exogenously applied COX substrate AA, which was in turn converted to prostanoids by the existing COX-1, could result in similar IL-8 production from these cells. Human ASM cells were cultured in the presence of AA (10 μM), and cell culture supernatants were collected at 0, 1, 2, 4, 8, 12, 16, 20, and 24 h. As shown in Figure 4,A, there was a time-dependent accumulation of PGE₂ that became significant after 1 h (p < 0.01) and peaked after 12 h of treatment (p < 0.001). When the cells were treated with AA at concentrations of 0.01, 0.1, 1.0, 10, and 100 μM for 16 h, a concentration-dependent increase in PGE₂ generation was also observed; this increase was significant from 0.01 μM (p < 0.05), and the highest concentration was obtained at 100 μM (p < 0.001, Fig. 4,B). In the meantime, AA also stimulated IL-8 production in a time- and concentration-dependent manner (Fig. 5, A and B) that was similar to that seen for PGE₂ generation; however, in the time course studies, the highest concentration of IL-8 was observed after 16 h of treatment, which was 4 h later than the time at which the highest concentration was observed for PGE₂.

To investigate whether COX products mediated the effect of AA on IL-8 production, we studied the effect of two of the nonselective COX inhibitors IND and FLU as well as other inhibitors such as NS-398, CHX, ACT, and DEX on AA-induced PGE₂ and IL-8 production from human ASM cells. AA (10 μM) markedly increased PGE₂ release, while both IND and FLU inhibited the increase in a strong and concentration-dependent fashion and abolished it at a concentration of 10 μM (Fig. 6,A). Conversely, NS-398, CHX, ACT, and DEX (all 1 μM) had no significant effect on the increase, as shown in Figure 6,B. These results suggest that only the constitutive COX-1 isoenzyme is involved in converting AA to prostanoids. Again, the same concentration range of IND and FLU also markedly inhibited AA-induced IL-8 production in a concentration-dependent fashion but did not abolish its effect (Fig. 7,A). In addition, NS-398, had no effect on AA induced IL-8 generation, similar to its ineffectiveness on PGE₂ release. However, CHX, ACT, and DEX suppressed IL-8 production (Fig. 7 B), suggesting that AA-induced IL-8 production is mediated by COX products from the COX-1 isoform and that the modulation of IL-8 expression by COX products is at the IL-8 gene transcription level.

To further clarify the role of COX products in BK-induced IL-8 production, we examined whether exogenously applied PGE₂ and calcium ionophore A23187, which has a similar effect to BK in causing free calcium increase and the release of endogenous AA, could result in similar IL-8 production. As shown in Figure 8, PGE₂ (1.0 μM) caused a sixfold increase in IL-8 production. A23187 was found to cause a concentration-dependent generation of PGE₂ after a 16-h incubation with human ASM cells. The maximum effect was observed with 10 μM of A23187, with which a 16-fold increase over basal PGE₂ was obtained (data not shown) and a nearly 18-fold increase of IL-8 production was also achieved (Fig. 8). Like that with AA, the IL-8 increase was strongly inhibited by IND, CHX, ACT, and DEX. These results consequently provide further evidence that COX products can cause IL-8 production from human ASM cells and are mainly responsible for the IL-8 production caused by BK.

To characterize the BK receptor(s) involved in BK-induced IL-8 production, we examined the effect of the selective BK B₁ receptor agonist desArg⁹-BK and the selective B₂ receptor agonist (Tyr(Me)⁸)-BK on this event. (Tyr(Me)⁸)-BK mimicked the effect of BK by causing IL-8 production in a concentration-dependent manner that was similar to that seen for BK; the increase was significant from 0.01 μM (p < 0.001), and maximum effect was observed at 1.0 μM (p < 0.001) (Fig. 9). In contrast, pretreating the cells with the B₁ receptor agonist desArg⁹-BK had only a weak effect on IL-8 formation at high concentrations; this effect was significant from 1.0 μM (p < 0.01) and reached a peak at 10 μM (p < 0.001) as compared with the control cells (Fig. 9). These results suggest that the B₂ receptor is responsible for the effect.

Pretreating human ASM cells with the selective B₂ receptor antagonist HOE-140 (1–100 μM) strongly antagonized BK (10 μM)-induced IL-8 production in a concentration-dependent fashion and abolished the effect of BK at 10 and 100 μM (Fig. 10); Over half (53%) of IL-8 production was inhibited by 0.1 μM HOE-140, and the IL-8 concentration was 138.67 ± 9.07 pg/ml. However, pretreatment with the B₁ receptor antagonist desArg⁹,(Leu⁸)-BK over the same concentration range did not show any significant effect (Fig. 10). Therefore, these data provide further evidence that B₂ receptors are responsible for mediating BK-induced IL-8 production from human ASM cells.

Cell viability after 16 h (some 24 h) of treatment with most of the chemicals used in this study was consistently >95% compared with cells that had been treated with the vehicles. However, the viability was reduced to ∼90% after the cells were treated with 100 μM of BK or 100 μM of AA for 16 h; therefore, 10 μM for both BK and AA was chosen for additional experiments in this study.

There are several novel findings in our study. This is the first study to show that human ASM cells are capable of IL-8 production in response to BK. It is also the first study to suggest that COX products play a critical role in BK-induced IL-8 production, and that COX products alone can stimulate IL-8 release. In addition, we characterized the BK receptor subtype and COX isoforms involved in these effects.

We found that stimulating human ASM cells with BK resulted in a time- and concentration-dependent release of IL-8 protein. ASM cells have long been studied largely from the perspective of target cells whose contractile and proliferational states are altered by local inflammatory events underlying the pathogenesis of asthma. Recently, however, a growing body of data has emerged to support the notion that human ASM cells have the potential to act as effector cells in perpetuating airway inflammation by expressing and secreting inflammatory products including prostanoids (18, 19, 20) and cytokines such as GM-CSF (15), IL-11, IL-6 (16), and RANTES (17). Our present observations that BK and AA cause IL-8 production from human ASM cells provide further evidence to support this notion. Thus, ASM may contribute directly to the recruitment of inflammatory cells such as neutrophils and eosinophils to the airways.

The most interesting finding of our study was that BK-induced IL-8 release was dependent upon COX products. This has not been previously reported in any cell system and may be relevant to the mechanism of action of BK in inflammatory diseases. Both the nonselective COX inhibitor IND and the selective COX-2 inhibitor NS-398 inhibited not only BK-induced prostanoid generation but also BK-induced IL-8 production in a similar manner. These findings were strengthened by experiments showing that the COX substrate AA alone also mimicked the effect of BK on prostanoid and IL-8 release; like that of BK, the effect of AA was also strongly inhibited by nonselective COX inhibitors as well as the protein synthesis inhibitors CHX and ACT and the steroid DEX. The effect of CHX, ACT, and DEX on BK- and AA-induced IL-8 release is likely to represent an effect on the transcription and translation of the IL-8 gene itself rather than an effect on COX-2 induction, since AA alone does not induce COX-2. The effectiveness of NS-398 on BK-induced prostanoid and IL-8 production and the ineffectiveness of NS-398 on AA-induced prostanoid and IL-8 production can be explained by the induction of COX-2 in BK- but not AA-treated human ASM cells. Since we have recently reported the details regarding BK-stimulated PGE₂ release and COX-2 induction in human ASM cells (19), Western blot results were not shown here. Our previous study showed that BK causes early PGE₂ release from constitutive COX-1 followed by later PGE₂ release due to COX-2 induction (19). We did not see a biphasic response with IL-8 production, which may reflect the time lag involvement in transcription and translation. The time allowance for the transcription and translation processes of IL-8 production could also explain why the peak for IL-8 release was 4 h later than that for PGE₂ release after AA treatment. The fact that IND produced a greater effect than NS-398 in the present study suggests that the prostanoids produced by constitutive COX-1 isoenzyme, in addition to those produced by inducible COX-2, play a role in BK-induced IL-8 production. Therefore, our findings provide the first direct evidence that BK induces IL-8 expression in human ASM cells and that COX products and COX-2 induction contribute to this process.

Although it has been well documented that cytokines can act as stimuli for IL-8 expression, the effect of other stimuli on IL-8 production is poorly understood. A limited number of reports have shown that calcium ionophore (31), β-adrenoceptor agonist isoproterenol (ISO) (32), and leukotriene (LT) B₄ (33) stimulate IL-8 release from various cell systems. Only one previous study has shown that BK can release IL-8; however, the study examined human decidua-derived cells and did not characterize the mechanism(s) involved (29). Our finding that BK can induce IL-8 production in airway cells may be of great relevance for asthma. In the present study, we used selective B₁ and B₂ receptor agonists and antagonists to show that BK causes IL-8 production via the B₂ receptor. This observation is consistent with the report that the BK-stimulated synthesis of IL-1β, IL-2, and IL-6 from isolated lung strips can be blocked by a B₂ receptor antagonist (27) and is also consistent with our previous study showing that the BK-mediated induction of COX-2 and the release of prostanoids from human ASM cells was mediated by the B₂ receptor (19). Therefore, B₂ receptor antagonists may play a role in controlling asthmatic airway inflammation.

Although the precise mechanisms of the effect of BK on IL-8 release are not fully understood, the results of our current study clearly demonstrate that prostanoid generation largely mediates the process. Prostanoids are generated by the oxidation of AA by COX. They are multifunctional mediators, but little is known about their effect on cytokine production. A recent report has shown that PGE₂ alone has little effect on IL-8 or GM-CSF but enhances IL-6 expression in human synovial fibroblasts; PGE₂ and iloprost (PGI₂ analogue) also enhance IL-8 production in IL-1α-stimulated cells (23). Our present study is the first to show that both endogenous (from AA) and exogenous prostanoids (PGE₂) directly cause IL-8 release from human ASM cells; their mechanism(s), however, remain to be investigated. It is known that the β-adrenoceptor agonist ISO enhances IL-8 release in airway epithelial cells via an increase in intracellular cAMP (32); our preliminary study has also shown increased IL-8 release after ISO stimulation of human ASM cells (our unpublished observations). Prostanoids (mainly PGE₂ and PGI₂) activate the PG EP2 and EP4 receptors that, like β-adrenoceptors, are coupled to adenylyl cyclase to increase intracellular cAMP production in human ASM cells (34). Thus, prostanoids share a similar receptor-mediated signal transduction system with β-adrenoceptor agonists. Consequently, it is reasonable to speculate that prostanoid-induced IL-8 release also occurs via the increase of intracellular cAMP. The same mechanism may also apply to IL-8 production caused by calcium ionophore (Ref. 31 and this study), because it also increases prostanoid generation. BK clearly alters airway function by several mechanisms that are in addition to the COX product-mediated effects on IL-8 seen in our paper. These include the stimulation of neural pathways and nitric oxide release. The relative role of these mechanisms remains to be elucidated.

Nonselective COX inhibitors were only partially effective on BK- (IND) and AA- (IND and FLU) induced IL-8 production (maximum inhibition of ∼80% for both inhibitors) despite completely blocking PGE₂ release from human ASM cells; this observation suggests that other products of AA may also be involved. The other major pathway for AA metabolism, apart from the COX pathway, is the generation of LTs via lipoxygenase. Inhibiting the COX pathway with COX inhibitors may result in the shunting of AA to the lipoxygenase pathway to produce more LTs. One of the major lipoxygenase products, LTB₄, possesses neutrophil chemotactic properties and has been shown to stimulate human polymorphonuclear leukocytes to synthesize and release IL-8 (33). Our preliminary investigation has also unveiled that exogenous LTB₄ alone induces IL-8 expression in a concentration-dependent manner with greater magnitude as compared with the same concentration of exogenous PGE₂ (our unpublished observations). Thus, it is likely that lipoxygenase products, to some extent, also mediate BK- and AA-induced IL-8 production in human ASM cells. Additional studies are needed to explore the lipoxygenase pathway of AA metabolism in human ASM cells under both resting and stimulated conditions.

In summary, this study examined the role of COX-2 induction and COX products in BK-induced IL-8 production in human ASM cells. Our results demonstrate that: 1) BK caused IL-8 production in cultured human ASM cells, and the effect was strongly inhibited by both nonselective and selective COX-2 inhibitors; 2) the COX substrate AA alone also stimulated IL-8 expression in a pattern that was similar to that seen for BK; 3) exogenously applied PGE₂ and calcium ionophore A23187 also caused IL-8 release; and 4) BK-induced IL-8 production was mimicked and abolished by a selective B₂ receptor agonist and antagonist, respectively. Collectively, these findings indicate that endogenous COX products, including those from COX-2 induction, are critically involved in mediating BK-induced IL-8 production in human ASM cells, and that human ASM may contribute directly to the recruitment of inflammatory cells in the airway and play a critical role in asthma pathogenesis. The mechanisms may be important in a number of other inflammatory diseases in which BK acts as a mediator.

We thank Colin Clelland for providing us with specimens of human trachea.