Asthma is a chronic inflammatory disease characterized by bronchial hyperreactivity and airway remodeling. Remodeling features are damage and shedding of airway epithelium, increased number of goblet cells, mucous gland hypertrophy, increased fibroblast/myofibroblast numbers, and increased airway smooth muscle mass and neovascularity [1–3]. Fibroblasts are thought to play a major role in the development of the structural changes in the airways by altering their phenotype from a quiescent cell into α-actin containing myofibroblasts that display enhanced proliferation, increased collagen and other extracellular matrix protein production, expression of adhesion molecules, and production of profibrotic and proinflammatory cytokines/chemokines and growth factors [1–4].

The early asthmatic response is typically initiated by activation of tissue-resident mast cells. The subsequent infiltration and activation of several additional inflammatory cells and notably that of the eosinophils account for the late phase of an asthma attack. Persistence of airway inflammation may explain the chronic disease. We and others have shown that mast cells and eosinophils are pivotal cells not only in initiating and maintaining airway inflammation in asthma but also in activation of fibroblasts to become myofibroblasts [5, 6]. The profibrotic effect of mast cells is due to several mediators, including histamine, tryptase, IL-6, nerve growth factor, eotaxin, TGF-β, TNF-α, and others [7–10].

Treatment of asthma usually includes short-acting β2-agonists for the relief of bronchoconstriction and, if needed, inhaled glucocorticosteroids to treat airway inflammation. Glucocorticosteroids have generally been considered to have antifibrotic activity based on data from in vitro studies carried out on fibroblasts from different anatomical origins. However, results obtained with lung fibroblasts have been inconclusive or contradictory because some have found that these drugs enhance fibroblast proliferation[11–14] and some have shown that they inhibit proliferation and production of profibrotic cytokines [12, 15]. Moreover, although very effective in reducing airway inflammation, the efficacy of glucocorticosteroids in attenuating airway fibrosis is still controversial [16–18]. Few studies have investigated the effect of short-acting β2-agonists on fibroblast proliferation and also with contrasting results [19, 20]. In this study, we were interested to examine the effect of salbutamol as a paradigm of short-acting β2-agonists and dexamethasone and as a paradigm of glucocorticosteroids and their combination on human fetal lung and adult bronchial fibroblast proliferation by themselves and in the presence of mast cells or eotaxin as stimulators.

Materials and methods

Fibroblasts

The human fetal lung fibroblast line MRC-5 (American Type Culture Collection, Manassas, VA) was cultured in Dulbecco's modified Eagle's medium (DMEM)/10% fetal calf serum (FCS) (DMEM supplemented with 10% heat-inactivated FCS, 2 mM L-glutamine, 100 U/mL penicillin, and 100 mg/mL streptomycin--all from Biological Industries, Beit Haemek, Israel) at 37°C, 5% CO2, and used for assays between the second and seventh passages. Human bronchial fibroblasts were obtained from endobronchial biopsy specimens of normal mucosa from 3 subjects (age, 54-64 yr; 1 female and 2 male subjects not receiving any medications) undergoing fiberoptic bronchoscopy (performed according to published guidelines) for assessment of contralateral localized lung cancer, as previously described [8]. Informed consent was obtained. Bronchial fibroblasts were expanded and cultured in DMEM/10% FCS and passaged weekly by Trypsin-EDTA (Biological Industries) treatment. They were used for assays between the second and fourth passages.

100 × Value post incubation with sonicates or drugs or mast cells or drugs and mast cells Value post incubation with medium alone .

Mast Cells and Mast Cell Sonicates

The human mast cell leukemia cell line HMC-1 (donated by Dr J. Butterfield, Mayo Clinic, Rochester, MN) was used as a source of mast cells [21]. HMC-1 cells were cultured in Iscove's medium (Biological Industries) supplemented with 10% FCS containing Fe [2], 50 U/mL penicillin, 0.05 mg/mL streptomycin, and 160 mL/L (1.2 mM) monothioglycerol. Cells were passaged twice a week.

HMC-1 cells were centrifuged (120 g, 5 min) and resuspended in 2 to 5 mL in DMEM/2% FCS (0.75-1:75 × 105/mL) and bath sonicated (2 min, 0°C; Heat Systems/Ultrasonic 380W, duty cycle 5 s; output power, 50%). Sonicates were then microcentrifuged for 5 minutes at 4°C, 120 g (Hettich Ultracentrifuge; Hettich Zentrifugen, Tuttlingen, Germany) to remove cell debris. The debris-free sonicate supernatant ("sonicate") was collected into aliquots and stored at -80°C.

Fibroblast Proliferation Assay

Fibroblast proliferation was assessed using the [3H]-Thymidine incorporation assay. Fibroblasts were plated in 96-well plates (3-4 × 103 cells per well in 200 μL DMEM/10% FCS). After incubation for 24 hours, the wells were rinsed twice with DMEM/2% FCS, and dexamethasone (Sigma Chemicals, St Louis, MO) (stock solution 10-2 M in ethanol) or salbutamol (stock solution 10-4 M in phosphate buffered saline; Sigma), at different concentrations or mast cell sonicates (15 × 103), or eotaxin/CCL11 (500 ng/mL) (R&D, Minneapolis, MN) was added in 200 μL DMEM/2% FCS in 3 to 4 wells for each treatment. Treatment with TGF-β (10 ng/mL) (R&D) was used as a positive control. After an additional incubation for 24 hours, 1 μL of [3H] Thymidine (2 or 5 mC/well) (DuPont NEN, Boston, MA) was added to each well, and the cells were incubated for an additional 24 hours. At the end of the incubation, cells were washed with cold phosphate buffered saline, fixed with methanol, and precipitated with 5% trichloroacetic acid (Merck, Darmstadt, Germany). The cells were lysed with 0.1% NaOH, transferred to scintillation vials, and counted in a β-counter (LKB 1211 Rackbeta; Pharmacia, Uppsala, Sweden). The results were obtained as counts per minute per well and are presented as the percentage of change (increase or decrease) in the fibroblast proliferation after the treatment compared with the cells that were incubated with medium alone:

IL-6 Release

To evaluate the presence of IL-6 released in the culture medium, an ELISA kit (PeptroTech, Rocky Hill, NJ) with sensitivity of 32 to 2000 pg/mL was used. For these experiments, fibroblasts were cultured in 24-well plates (3 × 103 cells per well in 200 μL DMEM/10% FCS) until confluent. At this point, the wells were rinsed twice with DMEM/5% FCS and dexamethasone, salbutamol, or forskolin (stock solution 5 mg/mL in ethanol; Sigma) at different concentrations was added in 200 μL DMEM/5% FCS. After an 18-hour incubation, supernatants were collected and kept frozen at -80°C until assayed.

Statistical Analysis

Each experiment was performed in at least triplicates for 3 or 4 times. The results are presented as the mean ± standard error of the mean and were analyzed by the unpaired Student t test or analysis of variance test as appropriate for each case. P values of < 0.05 were considered significant.

Results

Dexamethasone and Salbutamol Increase Fetal Lung Fibroblast Proliferation Alone and When Stimulated by Mast Cells

Fetal lung fibroblasts were incubated for 48 hours with different concentrations of dexamethasone or salbutamol, and fibroblast proliferation was monitored by [3H]-Thymidine incorporation (Figure 1). As shown in Figure 1A, the addition of dexamethasone caused upregulation in fibroblast proliferation in the range of concentration between 10-8 and 10-4 M in a bell-shaped manner. The maximal effect was achieved with 10-6 M (increase in proliferation, 350 ± 55%) (P < 0.0001) and was higher than that obtained with the addition of TGF-β. Addition of salbutamol (Figure 1B) also induced increased fibroblast proliferation, although to a lesser extent than dexamethasone. This increase was maximal at 10-9 M (187 ± 31% at 10-8 M) and was similar at 10-8 and 10-7 M. When dexamethasone and salbutamol were added together at the same concentration (10-5-10-8 M), the 2 drugs significantly increased fibroblast proliferation in a synergistic fashion (P < 0.0001), with the maximal increase of 432 ± 30% at 10-7 M (Figure 1C).

Figure 1
figure 1

The influence of dexamethasone and salbutamol on fetal lung fibroblast proliferation. Fibroblasts were incubated for 48 hours with medium alone or with either dexamethasone (10-4-10-8 M) (A) or salbutamol (10-6-10-10 M) (B) or dexamethasone and salbutamol (10-5-10-8 M) (C). TGF-β (10 ng/mL) was used as a positive control. Proliferation was assessed by [3H]-Thymidine incorporation, as detailed in Materials and Methods. Data are the mean ± standard error of the mean of 4 experiments performed in quadruplicates. In medium alone, counts per minute of [3H]-Thymidine incorporation were 474 ± 5. *P < 0.05; **P < 0.01; ***P < 0.0001.

Full size image

Next, we assessed the effect of the drugs on mast cell-induced fibroblast proliferation. Fetal lung fibroblasts were preincubated for 30 minutes with either dexamethasone or salbutamol, and then, mast cell sonicate (15 × 103 cells per well) was added for further 48 hours. As depicted in Figure 2, addition of mast cell sonicates increased fibroblast proliferation by 150 ± 26% (P < 0.05). Preincubation of dexamethasone at different concentrations further enhanced this effect (Figure 2A). The maximal increase in proliferation was observed with 10-8 M dexamethasone (349 ± 43%) (P < 0.0001). A similar result was obtained when the fibroblasts were incubated with different concentrations of salbutamol and mast cell sonicates, that is, salbutamol significantly increased fibroblast proliferation over the effect of the mast cell sonicates alone. The maximal increase in proliferation was observed with 10-7M salbutamol and was 348 ± 39% more than the control (P < 0.0001) (Figure 2B). In experiments performed with mast cell sonicate concentrations higher than 15 × 103 cells per well (ie, with 25 × 103 and 35 × 103 cells per well), still a significant increase in fibroblast proliferation (P < 0.0001) was observed with both dexamethasone (10-6-10-9 M) and salbutamol (10-5-10-8 M) preincubation.

Figure 2
figure 2

The influence of dexamethasone and salbutamol on mast cell-induced fetal lung fibroblast proliferation. Fibroblasts were preincubated for 30 minutes with either dexamethasone (10-6-10-9 M) (A) or salbutamol (10-5-10-8 M) (B). Mast cell (HMC-1) sonicates (15 × 103 cells per well) were then added, and cultures were incubated for further 48 hours. In the control, the fibroblasts were incubated in DMEM 2% alone or with mast cell sonicates alone. Proliferation was evaluated as described in the legend to Figure 1. Data are the mean ± standard error of the mean of 5 experiments performed in triplicates. In medium alone, counts per minute of [3H]-Thymidine incorporation were 219 ± 3. **P < 0.01; ***P < 0.0001.

Full size image

In Table 1, the results obtained with 10-8 M dexamethasone and 10-7 M salbutamol are depicted. Dexamethasone at 10-8 M enhanced significantly by 387 ± 41% the fibroblast proliferation when preincubated with fibroblasts, thereafter receiving as a trigger 35 × 103 mast cell sonicate per well with no significant increase, however, from lower mast cell numbers. On the other hand, salbutamol increased in a dose-response fashion the fibroblast proliferation with increasing numbers of mast cell sonicates to achieve a maximum of 566 ± 76% enhancement with 35 × 103 mast cell per well at a concentration of 10-7 M.

Table 1 Influence of Mast Cell Numbers on Dexamethasone and Salbutamol Increase in Fibroblast Proliferation
Full size table

Dexamethasone and Salbutamol Increase Human Bronchial Fibroblast Proliferation Alone and When Stimulated by Either Mast Cells or Eotaxin

When dexamethasone and salbutamol were added to human bronchial fibroblast subconfluent monolayers (Figures 3A, B), both drugs caused an increase in fibroblast proliferation (P < 0.05). Dexamethasone displayed the maximal increase at 10-8 M (174 ± 17%) and salbutamol at 10-6 M (145 ± 16%). Addition of mast cell sonicates (25 × 103 cells per well) alike increased fibroblast proliferation as expected. Proliferation was further increased when cultures were preincubated with either dexamethasone or salbutamol (Figure 4A). Cultures preincubated with dexamethasone displayed a significant increase in fibroblast proliferation compared with control and cultures incubated with sonicates alone. The maximal increase was obtained with 10-7 M dexamethasone and was 795 ± 90% (P < 0.0001). Salbutamol also increased the proliferative effect of mast cell sonicates (Figure 4B). Interestingly, in this case, the influence of salbutamol was stronger than that of dexamethasone at the same concentration, and the maximal effect was 1069 ± 96% at 10-7 M.

Figure 3
figure 3

The influence of dexamethasone and salbutamol on bronchial fibroblast proliferation. Fibroblasts were incubated for 48 hours with medium alone or with dexamethasone (10-7-10-9 M) (A) or with salbutamol (10-6-10-8 M) (B). Data are the mean ± standard error of the mean of 2 experiments performed in quadruplicates. In medium alone, counts per minute of [3H]-Thymidine incorporation were 270 ± 2. *P < 0.05.

Full size image
Figure 4
figure 4

The influence of dexamethasone and salbutamol on mast cell-induced bronchial fibroblast proliferation. Fibroblasts were preincubated for 30 minutes with either dexamethasone (10-7-10-9 M) (A) or salbutamol (10-6-10-8 M) (B). Mast cell (HMC-1) sonicates (15 × 103 cells per well) were then added, and cultures were incubated for further 48 hours. In the control, the fibroblasts were incubated in medium alone or with mast cell sonicates alone. Data are the mean ± standard error of the mean of 2 experiments performed in quadruplicates. In medium alone, counts per minute of [H3]-Thymidine incorporation were 270 ± 2. ***P < 0.0001.

Full size image

To check whether salbutamol enhances fibroblast proliferation when they are exposed to a defined profibrogenic compound, eotaxin was added to cultures preincubated with the drug for 48 hours. Eotaxin by itself increased fibroblast proliferation by 146 ± 9%. When together with salbutamol (10-5-10-8 M), a further significant increase was detected, that is, with a maximum of 261 ± 18% (P < 0.0001) reached with salbutamol concentration of 10-6 M (Figure 5).

Figure 5
figure 5

The influence of salbutamol on eotaxin-induced bronchial fibroblast proliferation. Fibroblasts were preincubated with salbutamol (10-5-10-8 M) for 30 minutes. Eotaxin (500 ng/mL) was then added to the wells for an additional 48 hours. In the control, the fibroblasts were incubated with medium alone or with eotaxin alone. In medium alone, counts per minute of [H3]-Thymidine incorporation were 330 ± 5. Data are the mean ± standard error of the mean of 5 experiments performed in triplicates. *P < 0.05; **P < 0.01; ***P < 0.0001.

Full size image

Dexamethasone and Salbutamol Influence Differently IL-6 Production From Fetal Lung Fibroblasts

As shown in Figure 6A, dexamethasone decreased in a dose-dependent fashion IL-6 production by confluent fetal lung fibroblasts (148% at 10-5 M; P < 0.005). Salbutamol in contrast enhanced this proinflammatory cytokine production also significantly with the maximal effect detected at 10-5 M of an increase of 129% (P < 0.01) (Figure 6B).

Figure 6
figure 6

The influence of dexamethasone and salbutamol on IL-6 production by fetal lung fibroblasts. Fibroblasts were incubated for 18 hours with medium alone or with medium with either dexamethasone (10-5-10-8 M) (A) or salbutamol (10-5-10-8 M) (B). IL-6 production was evaluated in the culture supernatants, as described in Materials and Methods. Data are the mean ± standard error of the mean of 2 experiments performed in quadruplicates.

Full size image

Interestingly, when the 2 drugs were added together at 10-5 and 10-6 M, the inhibiting effect of dexamethasone was slightly decreased by the addition of salbutamol (not shown). In a series of experiments (Table 2), salbutamol and forskolin were compared for their effects on IL-6 production from fetal lung fibroblasts. Similar increases were detected in IL-6 production relative to medium incubated wells. In the presence of salbutamol at 10-7 and 10-6 M, the increase in IL-6 found in the culture medium was 138 ± 6%* and 144 ± 4%*, respectively, and with forskolin at 10-6 and 10-5 M, it was 149 ± 4%* and 151 ± 4%*, respectively (n = 3, *P < 0.05).

Table 2 Influence of Dexamethasone, Salbutamol, and Forskolin on IL-6 Release From Fetal Lung Fibroblasts
Full size table

Discussion

It is well accepted that there is an increase in the number of fibroblasts in the airways of patients with asthma that correlates with thickness of lamina reticularis and disease severity. Moreover, fibroblast activation and differentiation to myofibroblasts are also evident [1–4].

In the present study, we aimed to investigate the in vitro effect of glucocorticosteroids and short-acting β2-agonists widely used as first-line antiasthmatic drugs on human lung fibroblast proliferation and IL-6 production. We specifically choose to evaluate fibroblast proliferation because this is the first hallmark of fibrosis taking place. IL-6 was selected among a plethora of proinflammatory profibrotic cytokines produced by the fibroblast[22] that mainly influences the inflammatory response [23, 24].

We found that dexamethasone and salbutamol alone and in combination increase both human fetal lung and human bronchial fibroblast proliferation. Moreover, we demonstrate for the first time that when the fibroblasts are preincubated with either drug and then with an activator such as mast cell sonicate or eotaxin, fibroblast proliferation is further increased. In contrast, the production of the proinflammatory and profibrotic cytokine IL-6 by confluent fibroblast monolayers was found to be decreased by dexamethasone, while it was increased in the presence of salbutamol. When the 2 drugs were added together, dexamethasone was still able to slightly decrease IL-6 production.

We have previously shown that mast cells enhance proliferation of human lung fibroblasts, an effect also seen with histamine and to a lesser extent with tryptase [7]. In a more recent study, we have also defined the profibrotic properties of eotaxin/CCL11, [8] a central mediator of asthma that can also be produced by mast cells [25].

In that study, we did not find significant differences between the human fetal lung fibroblast line and human bronchial fibroblasts regarding their proliferative responses to eotaxin/CCL11. This was explained by the fact that they were found to express similar levels of surface CCR3. Similarly, in the present study, we did not observe significant differences in the response of the fetal lung and the bronchial fibroblasts toward the proliferative effects of either the drugs or the stimulants. This is an interesting observation because fibroblasts from cell lines and primary cultures might behave differently. The fact that we did not detect major differences between these 2 sources of fibroblasts might be due to the fact that primary cells did not need to be cultured and subcultured for extensive time and when used were not senescent.

Many studies have been carried out to investigate the effects of glucorticosteroids on several sources of fibroblasts such as human fetal lines, or primary human lung, asthmatics endobronchial biopsies, human nasal polyps derived or mouse 3T3 and rat lung and even adipose tissue derived providing sometimes similar and sometimes contrasting results. We opted for the most commonly used source of human lung fibroblasts, the cell line MRC-5 and primary human bronchial fibroblasts.

Regarding our data, it is interesting to note that generally in the absence of the mast cell sonicate, dexamethasone effect was always greater than that of salbutamol. In this work, we showed for the first time that salbutamol caused a further increase in fibroblast proliferation in a dose-dependent fashion in fibroblasts treated with eotaxin. This proliferative event is probably carried out by the interaction of eotaxin with its receptor and by salbutamol with β2-receptors.

Glucocorticosteroids in general and specifically dexamethasone are known to decrease gene expression of proinflammatory cytokines. Several studies have found that dexamethasone decreases cytokine secretion by fibroblasts, among them IL-6 [26]. The results of our study correlate with these findings. But also this decreasing effect is not universally detected and depends on the exposure duration and the cytokine under study [27]. Studies that tested the effect of β2-agonists on IL-6 secretion on different cell types found that these drugs enhanced IL-6 secretion [28]. As far as we know, the effect of β2-agonists on IL-6 secretion from lung fibroblasts has not yet been studied. We found that salbutamol increased slightly IL-6 production. Cyclic adenosine monophosphate (cAMP) is known as an IL-6 secretion enhancer (airway smooth muscle cells, 3T3 fibroblasts, and gingival fibroblasts), even though it has been reported that lung fibroblast cell line had an inverse relationship between the amount of cAMP and the amount of IL-6 secreted from the cell [29].

In our study, salbutamol and forskolin that increase cAMP production, both increased IL-6 secretion from the fibroblasts. These findings strengthen the assumption that the mechanism through which salbutamol affects the increase in IL-6 secretion is cAMP dependent. Interestingly, when dexamethasone and salbutamol were added together, the decrease in IL-6 production was smaller than the one obtained in the presence of dexamethasone alone.

Many mediators, including growth factors, IL-6, and other cytokines, are involved in fibroblast proliferation regulation. In a previous study, we found that when added to fetal lung fibroblasts, IL-6 increased their proliferation up to 140% (data not published). Therefore, we might speculate that salbutamol can influence lung fibroblast proliferation by causing IL-6 production. The effect of β2-agonists in general and specifically of salbutamol on fibroblast proliferation has almost not been studied. Agarwal and Glasel described the effect of treatment with opioid and β-adrenergic agonists on cell proliferation [30]. They found that although both receptors increase cAMP production, opioid agonists lowered the level of proliferation, while a β2-agonist (procaterol) increased cell proliferation. Silvestri et al[20] found that salmeterol decreased fetal lung line fibroblast proliferation induced by bFGF. The source of the cells is indeed similar to the cells in our study, but salmeterol is a long-acting β2-agonist as opposed to salbutamol [20].

Another important difference is that the study by Silvestri et al, [20] and other studies, tested the drug's effect on proliferation of cells that were incubated with a growth factor (in this case basic fibroblast growth factor). It is difficult to determine whether the drug's effect is independent of the tested growth factor and whether this can be compared with our research in which salbutamol's effect on fibroblasts was assessed either without additional treatment or in the presence of mast cells or eotaxin.

It is difficult to speculate how the dexamethasone carries out its profibrogenic response. Several mechanisms have been suggested regarding the fashion in which the glucocorticosteroids enhance fibroblast proliferation. For example, they can either enhance or decrease the activity of different growth factors (among them, insulin growth factor-1 and fibroblast growth factor) and their receptors (ie, platelet derived growth factor[11]) or their production (ie, connective tissue growth factor[31]). The time of drug addition to the cultures can make a difference. When it is added after the cells reached confluence, it causes enhancement of proliferation, while drug administration during the exponential growth phase usually causes a decrease [32]. In our protocol, we added dexamethasone to subconfluent monolayer and still it had a profibrogenic effect.

It is interesting to point out that the literature on the effect of drug combination on fibroblast proliferation presents a complex picture. Many researches describe an additive and a synergistic effect of glucocorticosteroids and β2-agonists. On the other hand, other studies have found classes of the two drugs to have antagonistic effects. Peters et al[33] reported that β2-agonists, among them salbutamol, have the ability to lower the steroid effect on gene transcription by inhibiting the binding to the glucocorticosteroid responsive element on the DNA. On the other hand, salbutamol and salmeterol have been described as having the ability to activate the receptor by a mechanism related to cAMP increase [34].

Our findings obtained in vitro show that whatever the underlying mechanism(s) is, the drugs have a direct fibrogenic effect and a synergistic effect with central players, such as mast cell-derived mediators in the asthmatic process. If these in vitro observations are found to be true also in the in vivo complexity of the asthmatic patient with airway remodeling and fixed airway obstruction, it is then possible to speculate that this may not only be part of the natural course of the disease but a complication of the accepted drug treatment and a product of the drugs' interactions with the inflammation mediators. Therefore, if the profibrogenic effect of glucocorticosteroids and β2-agonists is reconfirmed in vivo, the combination of the remodeling and the drugs' effect can be dangerous. In conclusion, more in vivo studies are warranted to clarify what the risks and benefits are of monotherapy versus combination of inhaled glucocorticosteroids and short- and long-acting β2-agonists on airway remodeling in asthma.

Note

Partially funded by a research grant from the Aimwell Charitable Trust to F. Levi-Schaffer.