Published ahead of print on September 3, 2004, doi:10.1165/rcmb.2003-0331OC
© 2004 American Thoracic Society DOI: 10.1165/rcmb.2003-0331OC Endothelin-1 Inhibits Mucin Secretion from Ovine Airway Epithelial Goblet CellsSchool of Pharmacy and Biomolecular Sciences, University of Brighton, Brighton; Pfizer Global Research and Development, Sandwich, Kent; Respiratory Cell and Molecular Biology Division, University of Southampton, Southampton, United Kingdom; Department of Anatomy II: Experimental Morphology, University of Hamburg, Hamburg, Germany Address correspondence to: Dr. Michael I. Lethem, School of Pharmacy and Biomolecular Sciences, University of Brighton, Brighton BN2 4GJ, UK. E-mail: m.i.lethem{at}brighton.ac.uk
Mucus hypersecretion is a feature of several respiratory diseases and frequently leads to obstruction of small airways where the principal source of mucous glycoproteins (mucins), the major macromolecular constituents of mucus, are goblet cells. Hence, inhibition of mucin secretion from these cells may be clinically beneficial. In this study, we have developed a lectin-based assay for mucin secretion from ovine airway goblet cells and used this assay to investigate the regulation of these cells by endothelin (ET)-1. ET-1 inhibited baseline mucin secretion (maximum inhibition: 60.3 ± 4.2%, 50% inhibitory concentration: 0.8 ± 0.17 nM). This response was abolished by the ETA antagonist, BQ-123 (1 µM), but not by the ETB antagonist, BQ-788 (1 µM). ET-1 (1 µM) did not affect mucin secretion stimulated by ATP (100 µM) but secretion in response to ATP (10 µM) was inhibited by 63.3 ± 11.8%. This response could be eliminated by BQ-123, but not by BQ-788. Radioligand binding and immunohistochemistry indicated the expression of both ETA- and ETB-receptors on the epithelium. In summary, ET-1, acting via ETA-receptors, inhibits baseline and ATP-stimulated mucin secretion from ovine airway goblet cells. This represents the first report of a physiologic mechanism for inhibiting airway goblet cell mucin secretion; an understanding of this mechanism may provide opportunities for the treatment of obstructive airways disease.
Abbreviations: maximum binding, Bmax • calcium, Ca2+ • Chinese hamster ovary, CHO • distilled water, dH2O • enzyme-linked lectin-binding assay, ELLA • endothelin, ET • Helix pomatia agglutinin, HPA • dissociation constant, KD • minimal essential medium, MEM • mucous glycoprotein, mucin • phosphate-buffered saline, PBS • phospholipase C, PLC • sarafatoxin S6b, SRTX6b • sarafatoxin S6c, SRTX6c
Mucus secretion into the airway is a central component in the process of mucociliary clearance; the respiratory defense mechanism that entraps inhaled foreign material in mucus and removes it from the lung by the action of beating cilia. The physical properties of mucus are principally determined by the mucous glycoproteins (mucins), which are derived from two principle sources: mucous cells of the submucosal glands, and goblet cells situated in the epithelium lining the airways (1). The presence of these multiple sources of mucin in the respiratory tract has hindered investigations of the regulation of secretion from specific cell types, in particular goblet cells. Studies involving the regulation of secretion from goblet cells are of particular importance because goblet cells are the sole source of mucin in the small airways, which are the principle sites of obstruction in airway diseases associated with hypersecretion (2). Hence, the possibility of inhibiting the level of mucin secretion from airway epithelial goblet cells in hypersecretory diseases has therapeutic potential. However, to date, no physiologic mechanism for inhibiting mucin secretion from goblet cells has been identified, although previous studies have suggested that endothelin (ET)-1 can inhibit mucin secretion from cat and ferret tracheal submucosal glands (3, 4). An increasing amount of literature from in vitro and in vivo studies indicates that ET-1 mediates an array of effects in the lung, many of which have suggested a pathophysiologic role (5). Two distinct receptors, designated ETA and ETB, have been cloned and characterized, and both are widely distributed in the mammalian lung (6). Activation of ETA-receptors in human airways is responsible for ET-1–induced prostanoid release (7) and smooth muscle proliferation (8), whereas stimulation of ETB-receptors potentiates cholinergic nerve-induced contraction (9). The effect of ET-1 on mucin secretion from airway goblet cells is unknown. We therefore investigated, in the present study, the effect of ET-1 on baseline and ATP-stimulated mucin secretion from goblet cells in ovine tracheal epithelial explants and the expression of ET-receptors in ovine tracheal epithelium.
Development of Mucin Assay In order to assess the mucin secretory activity of airway epithelial explants, a lectin-based assay for ovine goblet cell mucin was developed. To identify lectins with specificity for goblet cells in the epithelium, ovine tracheal tissue was obtained from the abattoir and tissues, or epithelial explants isolated therefrom, and fixed in 4% paraformaldehyde. Sections were cut from the paraffin-embedded fixed tissues and screened against a panel of 18 peroxidase-labeled lectins using standard immunohistochemical techniques, and the epithelial staining pattern was assessed. A number of lectins showed preferential staining of goblet cells in the epithelium, and of these, Helix pomatia agglutinin (HPA) was chosen for incorporation into a microplate-based mucin assay (see below), which was then used to monitor the mucin secretory activity of airway epithelial explants in subsequent studies. The nature of the HPA-positive material released from airway epithelial explants was characterized further by cesium chloride density gradient ultracentrifugation analysis of material released from explants in response to the known mucin secretagogue, ATP (100 µM). Airway epithelial explants were prepared from ovine trachea as described below and, after periods for equilibration and assessment of baseline mucin secretion, were exposed to ATP (100 µM) in Ham's F12 medium for 30 min, with the medium being collected and replaced with fresh Ham's F12 containing ATP every 10 min. The ATP-stimulated samples from eight separate explants were pooled and subjected to isopycnic density gradient ultracentrifugation, according to the method of Carlstedt and colleagues (10). Briefly, guanidine hydrochloride was added to give a final concentration of 4 M and cesium chloride was added to give a density of 1.39 g/ml, and the mixture centrifuged at 120,000 x g for 72 h at 15°C. The resulting density gradient was fractionated into 1 ml fractions, which were then monitored for density by weighing and for HPA-positive material by lectin-based assay, as described below.
Epithelial Explant Preparation
Mucin Secretion Studies
Quantification of Mucin Secretion
Radioligand Binding Binding of [125I]–ET-1 (specific activity of 2,200 Ci/mmol) (Amersham Pharmacia Biotechnologies Ltd., Little Chalfont, UK), and the ETA-receptor–selective ligand [125I]–PD151242 (specific activity of 2,200 Ci/mmol) (Amersham Pharmacia Biotechnologies Ltd.) to membranes prepared from ovine tracheal epithelium was measured in triplicate after 2.5 h incubation at 23°C in binding buffer (10 mM ethylenediaminetetraacetic acid [GIBCO Laboratories], 50 µg/ml bacitracin [Sigma Chemical Co.], and 30 µM phenylmethylsulfonyl fluoride [Sigma Chemical Co.]). Binding was performed in 96-well filter binding plates (0.66 µm) (Millipore UK Ltd., Washington, UK). Total binding was assessed by measuring binding of the above ligands to membrane preparations (15 µg/well, 50 µl) in binding buffer only, and nonspecific binding was assessed in an excess of the nonselective ETA and ETB-receptor ligand UK240441 (10 µM) (kindly provided by Tony Chuck from Pfizer Global Research and Development, Sandwich, Kent, UK) in binding buffer, and competitive binding studies were performed in the presence of the appropriate competing ligand (20 µl) in binding buffer. Reactions were started by the addition of [125I]–ET-1 (125 fM–2 nM, 30 µl) or [125I]–PD151242 (8 pM–8 nM, 30 µl), with the exception of competitive binding studies, where the membrane homogenates were pretreated for 20 min with varying concentrations of the ETA-receptor–selective ligand BQ-123 or the ETB-receptor–selective BQ-788. After incubation, the reactions were terminated by rapid filtration using a vacuum filtration manifold (Millipore UK Ltd.). To separate membrane-bound radioactivity from free ligand, filters were washed four times in rapid succession with 200 µl of ice-cold Tris buffer and left to dry at 40°C. ‘Microscint’ (50 µl) (Packard Instrument Co., Cambridge, UK) was added to each well and bound radioactivity was measured. Typically, nonspecific binding was between 10–15% of total binding, and at the protein concentrations used [125I]–ET-1 and [125I]–PD151242 binding reached equilibrium by 2.5 h at 23°C. All binding experiments were conducted using three membrane preparations, from 3 different animals, pooled together.
Immunohistochemical Localization of ETA- and ETB-Receptors To assess the specificity of the ET-receptor antibodies, Chinese hamster ovary (CHO) cells, transfected to express either human ETA- or ETB-receptors (kindly provided by Tony Chuck from Pfizer Global Research and Development), were used as controls for immunohistochemical staining. CHO cells were cultured in Ham's F12 containing 10% fetal calf serum, 2 mM L-glutamine, and 500 µg/ml gentamicin (Sigma Chemical Co.). After passage, cells were resuspended in PBS and between 1.0 x 10–3 and 1.0 x 10–4 cells/ml were centrifuged onto Cytospin slides (Vector Laboratories Inc.). Slides were fixed in ice-cold 4% paraformaldehyde in dH2O, for 5 min at 4°C, washed 4 times in PBS, and air-dried. Slides were then probed with anti-ETA- and anti-ETB-receptor antibodies, as described above.
Solutions and Drugs
Statistical Analysis Results from saturation binding experiments were analyzed with the iterative curve-fitting program LIGAND and are presented as means ± SEM. Data for competition experiments were analyzed using GraphPad Prism (GraphPad, San Diego, CA), comparing one-site and two-site models, and Hill coefficient (nH) values were determined for all experiments. Differences among values were tested for significance by comparing the residual variance using an F test and a significance level of P < 0.05.
HPA Staining and Mucin Assay Staining of sheep tracheal sections with a panel of lectins revealed that a number of these showed a degree of selectivity for goblet cells in the epithelium. In particular, Helix pomatia agglutinin stained goblet cells without staining the ciliated border of the epithelium (Figure 1A). Some areas of discreet staining on the surface of the tissue were apparent; however, this does not appear to have been associated with the cilia, and may reflect staining of adherent mucus on the surface of the tissue. Using a purified human respiratory mucin standard, HPA was incorporated into an ELLA, the standard curve for which, when plotted as the log of absorbance versus the log of mucin concentration, was linear in a range from < 10 ng/ml to > 250 ng/ml (Figure 1C). To assess whether the HPA-binding assay could be used to monitor the secretory activity of ovine airway epithelial explants, the response of explants to the mucin secretagogue ATP was investigated. After equilibration, the explants achieved a steady baseline of 11.5 ± 6.2 ng human mucin equivalent per 10 min incubation period. Exposure of the explants to ATP (100 µM) resulted in an increase in HPA-positive material to 277 ± 42.7% of baseline within 10 min of application, a response which returned to near baseline levels over the period of the 30 min exposure (Figure 1D). Isopycnic density gradient ultracentrifugation of the HPA-positive material produced by explants in response to ATP indicated that this material was a single band in the density gradient with a peak buoyant density of 1.38–1.53 g/ml (Figure 1E).
ET-1 Inhibits Baseline and Stimulated Mucin Secretion ET-1 (1 µM) inhibited baseline mucin secretion by 63.5 ± 9.6% within 10 min of exposure (P < 0.01) (Figure 2A), an effect that was sustained throughout the 40 min ET-1 exposure period. Over the concentration range 1 pM–1 µM, the inhibitory effect of ET-1 on baseline mucin secretion was concentration-dependent, with a 50% inhibitory concentration of 0.8 ± 0.17 nM (Figure 2B) and a maximum inhibition at 1 µM ET-1, which reduced baseline mucin secretion by more than 60% to 39.7 ± 4.2% of the level seen before ET-1 application. In two further experiments, in which the ET-1 concentration was increased to 10 µM, no further inhibition was observed (data not shown).
In order to investigate the role of ET-receptors in the effect of 1 µM ET-1 on mucin secretion, the effects of the ETA antagonist BQ-123 and the ETB antagonist BQ-788 (both at a concentration of 1 µM) were studied. BQ-123 alone had no significant effect on baseline mucin secretion (Figure 3A). However, BQ-123 abolished the ET-1–evoked inhibition of mucin secretion, although after a 20-min washout period, ET-1 alone inhibited baseline mucin secretion by 64.4 ± 4.4% (P < 0.01). Pre-exposure of the explants to the ETB-receptor antagonist BQ-788 had no significant effect on baseline mucin secretion (Figure 3B); however, in the presence of BQ-788, ET-1 still inhibited mucin secretion by 54.8 ± 8.1% (P < 0.01). After the 20-min washout period, the ability of ET-1 to inhibit baseline mucin secretion increased to 79.5 ± 8.2% (P < 0.01), a response that was significantly (P < 0.01) greater than that seen for the combination of ET-1 and BQ-788.
The role of ET-receptors in the regulation of baseline mucin secretion was investigated further with the ET-receptor agonists sarafatoxin S6b (SRTX6b) and sarafatoxin S6c (SRTX6c). The ETA-selective agonist, SRTX6b (100 nM) resulted in a 56.8 ± 5.1% inhibition of baseline mucin secretion (P < 0.01) (Figure 3C), a response similar to that observed with ET-1 (1 µM). This inhibitory effect was sustained for the 30-min agonist exposure period. However, the ETB-receptor–selective agonist, SRTX6c (100 nM), did not significantly affect baseline mucin secretion levels throughout the 30-min treatment period (Figure 3D). Thus, the ETA-receptor and not the ETB-receptor appears to mediate the inhibitory effect on mucin secretion. To determine whether ET-1 inhibited stimulated as well as baseline mucin secretion, the effect of ET-1 in combination with the mucin secretagogue, ATP, was studied. Exposure of the explants to ATP (100 µM) caused an increase in mucin secretion over and above the baseline level of 199.1 ± 45.1%, a response that was not significantly altered by the addition of ET-1 (1 µM) (Figure 4A). However, although stimulation of the explants with ATP alone at the lower concentration of 10 µM still resulted in a significant (P < 0.05) increase in mucin secretion of 77.3 ± 10.9%, the addition of ET-1 (1 µM) in the presence of ATP (10 µM) significantly reduced (P < 0.01) this response to an increase of only 28.3 + 9.9% (Figure 4B). The effect of ET-1 on mucin secretion stimulated by ATP (10 µM) was abolished by BQ-123 (1 µM), with the response to ATP rising to 134.6 ± 32.6%. This increase in the response to ATP in the presence of BQ-123 was, however, largely attributable to a single explant that exhibited an unusually robust response to ATP. In the presence of the selective ETB-receptor antagonist, BQ-788 (1 µM), ET-1 still significantly (P < 0.05) inhibited mucin secretion stimulated by ATP (10 µM) (Figure 4B).
Radioligand Binding Studies Binding of [125I]–ET-1 (nonselective ligand interacting with ETA and ETB-receptors) and [125I]–PD151242 (ETA-receptor–selective ligand) to membranes prepared from ovine tracheal epithelium was dependent on the membrane protein concentration as well as the incubation time (data not shown). Analysis of saturation binding data revealed that [125I]–ET-1 and [125I]–PD151242 bound with high affinity to membranes prepared from ovine tracheal epithelium, with apparent dissociation constants (KD) of 0.16 ± 0.09 nM and 0.66 ± 0.06 nM, respectively (Table 1). Over the concentration range tested, Hill slopes were close to unity, and a one-site fit was preferred to a two-site model, suggesting that both ligands bind to either a single population of receptors or to a heterogeneous population with equal affinity. Maximum binding (Bmax) values were 1,170 ± 120.5 and 280 ± 20.1 fmol/mg protein for [125I]–ET-1 and [125I]–PD151242, respectively, suggesting the presence of both ETA- and ETB-receptors.
The ETA-selective ligand BQ-123 (12.8 pM–25 µM) competed for [125I]–ET-1 binding in a monophasic manner, a one-site fit being preferred to a two-site model (Figure 5A). At the highest concentration tested (25 µM), BQ-123 competed for 58.1 ± 3.4% of the binding, and a KD value could not be calculated. Concomitant treatment with BQ-788 (10 pM–40 nM) completely inhibited the BQ-123–resistant component of [125I]–ET-1 binding (data not shown). The ETB–selective ligand BQ-788 (10 pM–62.5 nM) competed for almost 100% of [125I]–ET-1 binding (Figure 5B). However, inhibition curves were biphasic, indicating the presence of high-affinity, high-Bmax sites (KD, 0.82 ± 0.06 nM; Bmax, 696.5 ± 73.5 fmol mg-1 protein) and the lower-affinity, lower-Bmax sites (KD, 57.8 ± 7.08 nM; Bmax, 197.6 ± 23.4 fmol mg-1 protein). Approximately 72% of the sites were high affinity and 28% were low affinity. Pseudo–Hill coefficients were less than unity (0.42 ± 0.05) and a two-site fit was preferred to a one-site model.
Similar competition binding experiments were performed using the selective ETA-receptor ligand [125I]–PD151242. As expected, BQ-123 (64 pM-25 µM) competed for [125I]–PD151242 binding (KD, 0.89 ± 0.19 nM) in a monophasic manner, with close to 100% inhibition of binding (Figure 5C), and a one-site fit was preferred to a two-site fit. The ETB-selective ligand, BQ-778 (40 nM–10 µM), at the highest concentration competed for 60% of the ETA-selective [125I]–PD151242 binding, and KD value could not be calculated (Figure 5D).
Immunohistochemical Studies
A major limitation of studies of airway mucin secretion using whole tracheobronchial tissue has been the difficulty in distinguishing between the mucins secreted by goblet cells and submucosal glands. The results presented in this study indicate that the technique of isolating and explanting intact sheets of epithelium, as described in previous studies of mucin secretion from goblet cells in canine (11) and human tracheae (12), is also applicable to ovine airways. This explant preparation has the advantage over mucin-producing primary cultures of airway epithelium in that cellular differentiation occurs through the normal processes in the host organism. Furthermore, the unchanging morphology (11) during the isolation and explantation procedure suggests that tissue damage is minimal, increasing the likelihood that the explants accurately model the characteristics of the native tissue. To use the ovine airway epithelial explants for studies of mucin secretion, it was necessary to develop an assay that would selectively measure mucin production from the explants. This was achieved using an ELLA with HPA as the lectin of choice. The choice of HPA for the assay was based on the staining pattern observed in ovine tracheal tissues, its reactivity with a human mucin standard in a microtitre plate assay, its ability to detect changes in mucin production in explants treated with a known secretory stimulus, and the migration pattern of the HPA-positive material in a cesium chloride density gradient.
Staining of ovine airway epithelial tissues with HPA indicated that the lectin bound to goblet cells within the epithelium without binding to the ciliated border, as has been reported for a number of antimucin antibodies (12, 13). This pattern of staining is consistent with the lectin binding selectively to mucins, a finding that is supported by the ability of HPA to recognize a human respiratory mucin standard in the lectin-binding assay. Further evidence of lectins usefulness in measuring mucin secretion from explants can be found in the fact that it allowed the detection of changes in mucin secretion by explants in response to the known mucin secretagogue, ATP. Explants incubated with F12 alone showed a stable level of baseline mucin secretion; however, the absolute values for the amount of mucin produced by each explant was highly variable and ranged from 1.3–33.8 ng mucin per 10 min period. The reason for this variability is not clear, and could possibly reflect differences in the secretory activity of different explants, but could also be a reflection of differences in the reactivity of mucins from different explants with the HPA. Differences in the reactivity of different mucins with HPA should also be considered with respect to the use of the human mucin standard in the assay. It is quite possible that the HPA has a different sensitivity for the human mucin standard than for the ovine mucins produced by the explants; hence, care should be taken in interpreting the values for the absolute amounts of mucin produced by the explants. Irrespective of the levels of baseline mucin secretion observed, all explants responded similarly to stimulation by ATP with a transient increase of The present study has established that ET-1 dose-dependently inhibits baseline mucin secretion from ovine tracheal epithelial goblet cells (50% inhibitory concentration, 0.8 ± 0.17 nM; maximum inhibition, 60.3 ± 4.2%) and represents the first report of a physiologic inhibitor of airway goblet cell mucin secretion. Previous studies by Shimura and colleagues (3) investigating the effect of ET-1 on feline tracheal submucosal glands found that, although ET-1 stimulated mucin secretion in isolated glands, it inhibited mucin secretion in the presence of epithelial cells, suggesting the involvement of an epithelium-derived inhibitory factor. Furthermore, the facts that indomethacin could partially inhibit this effect and that ET-1 has been reported to induce the release of phospholipase A2 products (16), suggest that this factor may be an eicosanoid. It is not known from the present study whether ET-1 inhibits baseline mucin secretion from ovine tracheal goblet cells indirectly, via the release of epithelium-derived inhibitory factor(s), or through a direct action of ET-1 on mucin-secreting cells. In addition, the intracellular signaling pathways responsible for this action are not known. ET-receptors are coupled to a number of effectors to produce an extensive network of second messengers; for example, stimulation of ETA-receptors and/or ETB-receptors in various cell types have been linked to various intracellular pathways, including phospholipase C (PLC) (17), phospholipase A2 (16, 18), guanosine 3'5'-cyclic monophosphate (cGMP) (19), and phospholipase D (20). In addition, calcium (Ca2+) signaling appears to be an almost universal response to ET-induced receptor activation (4, 21, 22). However, previous studies suggesting that rises in intracellular Ca2+ stimulate mucin secretion from airway goblet cells (13) call the role of Ca2+ in ET-1–mediated inhibition of stimulated mucin secretion into question. ET-1 had no effect on mucin secretion in the presence of BQ-123, indicating that ETA-receptors mediate the inhibitory action of ET-1—a finding supported by the fact that the ETA-receptor–selective agonist SRTX6b inhibited mucin secretion from the ovine goblet cells to an extent similar to that observed with ET-1. BQ-788 was also capable of significantly reducing the effect of ET-1 on baseline mucin secretion, although to a much lesser extent than BQ-123. However, the role for ETB-receptors that this suggests is not supported by the studies with SRTX6c (Figure 3D), which indicated that activation of ETB-receptors had no effect on baseline mucin secretion. Although it is possible that this contradiction might be the result of an inability of SRTX6c to mediate an effect from ETB-receptors at this concentration, this seems unlikely, because previous studies, although in other species, have suggested that SRTX6c is active at ETB-receptor sites at the concentration used here (22). It seems more likely that the ability of BQ-788 to inhibit the action of ET-1 is the result of lack of selectivity for ETB-receptors. This hypothesis is supported by the radioligand binding studies in which BQ-788 competed for almost 100% of [125I]–ET-1 binding in a biphasic manner, and also competed with the ETA-selective ligand [125I]–PD151242 (23) (Figures 5B and 5D). The selectivity of BQ-788 has similarly been questioned in previous studies in rat lung, where the binding of [125I]–ET-1 to both high-affinity ETB-receptor sites and low-affinity ETA-receptor sites was inhibited by BQ-788 (24). Furthermore, the affinity of BQ-788 for ETB-receptors has been reported to differ between rat and human tissues (25), raising the possibility of species differences in the binding characteristics of BQ-788 to ET-receptors. Such species differences might explain the results obtained with ovine tissues in the present study. ET-1 (1 µM) did not affect the ability of the explants to respond to 100 µM ATP; however, although the lower concentration of 10 µM ATP was still able to elicit a significant increase in mucin secretion from the explants, ET-1 (1 µM) was able to inhibit this response. The reason for the difference in the effect of ET-1 at the two ATP concentrations is not clear. However, studies by Abdullah and colleagues (26) showed that ATP dose-dependently stimulated mucin secretion from the rat tracheal epithelial cell line, SPOC 1, with an apparent K0.5 of 4 µM, and maximum stimulatory response was achieved at 100 µM ATP. Similarly, Kim and Lee (14) concluded that ATP dose-dependently stimulated mucin secretion from hamster tracheal epithelial cells, with an apparent EC50 of 20 µM and near-maximal mucin stimulation at 100 µM ATP. It is possible, therefore, that the lack of an action of ET-1 on mucin secretion stimulated by ATP at 100 µM was due to the near maximal stimulus of mucin secretion at this concentration, a stimulus that overshadowed the inhibitory effect of ET-1. At the lower concentration of ATP, the stimulatory drive for secretion would have been reduced, thus unmasking the inhibitory effect of ET-1. The ability of ET-1 to inhibit stimulated mucin secretion is in agreement with previous studies in ferret submucosal glands, which indicated that ET-1 was capable of inhibiting methacholine- and phenylephrine-stimulated mucus secretion (4). In the present study, the inhibitory action of ET-1 on ATP-stimulated mucin secretion could be blocked by BQ-123 but not by BQ-788, indicating that the effect of ET-1 was mediated by ETA-receptors but not ETB-receptors. The mechanism by which ET-1 inhibits ATP-stimulated mucin secretion is not clear. In a number of systems, stimulation of ETA-receptors has previously been linked to activation of PLC and mobilization of intracellular Ca2+ (4, 21). However, the role of PLC in the effect of ET-1 on ATP-stimulated mucin secretion is questionable, as purinergic stimulation of goblet cell mucin secretion is also mediated, in part, by a PLC and Ca2+ pathway in a number of systems (27, 28). The radioligand binding studies reported here demonstrate that ovine tracheal epithelium expresses both ETA- and ETB-receptors. [125I]–ET-1 binds with high affinity to ovine tracheal epithelium. The KD value for [125I]–ET-1 (0.16 ± 0.09 nM) was in a range similar to that reported for other tissues (e.g., rat lung, 0.15 nM [24], and canine tracheal epithelial cells, 0.2nM [29]). The ETA-receptor antagonist, BQ-123, competed for almost 60% of [125I]–ET-1 binding and the BQ-123 insensitive component of [125I]–ET-1 binding was abolished by concomitant treatment with BQ-788, suggesting the presence of both ETA- and ETB-receptor sites. However, the ETB-receptor–selective ligand, BQ-788, competed for almost 100% of [125I]–ET-1 binding in a biphasic manner, indicating the presence of high-affinity (subnanomolar range) ETB-receptor sites (71.6% of sites) and lower affinity (nanomolar range) non–ETB-receptor sites. This receptor ratio is comparable with the saturation binding studies with [125I]–ET-1 and [125I]–PD151242, where the percentage difference between the Bmax values for the nonselective ([125I]–ET-1) and ETA-selective ([125I]–PD151242) ligands revealed a receptor ratio of 24:76 (ETA:ETB) (Table 1). However, in competition binding studies, the ETA-selective antagonist BQ-123 inhibited almost 60% of [125I]–ET-1 binding, and inhibition curves were monophasic, suggesting binding to only one receptor subtype. These results are difficult to explain without postulating the existence of additional receptor subtypes, or may alternatively be explained by the different agonist versus antagonist binding affinity profiles observed with G-protein–coupled receptors; where only agonist binding is affected by the G-protein–coupling state or the absence or presence of guanine nucleotides (30). It is also interesting to note that, although BQ-788 was able to almost completely inhibit [125I]–ET-1 binding in the binding assay, it was not able to block the effect of ET-1 on mucin secretion from the explants (Figure 3B). This apparent discrepancy is probably a reflection of the different systems used in these two assays, and also of the fact that in the mucin secretion studies, ET-1 and BQ-788 were present in equimolar concentrations, whereas in the ligand binding studies, BQ-788 was present in a large excess compared with the concentration of [125I]–ET-1. The ETA-selective linear tetrapeptide radioligand [125I]–PD-151242 (23) was used to characterize ETA-receptors in ovine tracheal epithelium. Saturation binding assays with [125I]–PD-151242 revealed a single population of high affinity ET-receptors: (KD = 0.66 ± 0.06 nM). The dissociation constant for [125I]–PD-151242 was in a range similar to other tissues that contain mainly ETA-receptors, such as human vasculature and heart (aorta [0.8 nM], coronary artery [0.5 nM], pulmonary artery [1.8 nM], and ventricle [1.1 nM]) (23). BQ-123 and BQ-788 competed for [125I]–PD151242 binding in a monophasic manner, but the former with high-nanomolar affinity and the latter with micromolar affinity. BQ-123 competed for almost 100% of [125I]–PD-151242 binding, and BQ-788 (1 µM) competed for almost 60% of ETA-selective [125I]–PD151242 binding, once again calling in to question the selectivity of the ETB-receptor antagonist in this preparation. Pharmacologic studies in animal tissues have suggested the presence of an atypical ETA-receptor (e.g., in rat kidney) (31). However, as BQ-123 competed for almost all of the binding of [125I]–PD151242 to ETA-receptors, and inhibition curves were monophasic, there is nothing to suggest that atypical ETA-receptors are present in ovine tracheal epithelium. Immunohistochemical studies indicated both ETA- and ETB-receptor like immunoreactivity throughout the epithelial layer. ETA-receptor–like immunoreactivity was expressed predominantly on the apical cell layer and on the basal cells of the epithelium, with lateral bands of immunoreactivity through the epithelial layer also being evident. Ninomiya and colleagues (32) found a similar expression pattern of ETA-receptors in basal cells from guinea pig tracheal epithelium and suggested that ET-1 activates ETA-receptors and stimulates the proliferation of the basal cells. This mitogenic effect is consistent with previous observations in porcine tracheal epithelial cells (33), as well as other cell types that express the ETA-receptor subtype (34). ETB-receptor–like immunoreactivity was expressed on cilia and along the apical surface of the epithelial layer. Similarly, in guinea pig trachea, expression of ETB-receptors was located on the cilia (32). The localization of ETB-receptors on cilia and ciliated cells suggests that ETB-receptors are important in controlling ciliary beat frequency in ovine trachea. ET-1 depresses tracheal mucus velocity in ovine airways (35); however, antagonist studies suggest that this response is mediated via ETA-receptors. Mucus velocity is dependent on other factors in addition to ciliary beat frequency (e.g., the composition of airway surface liquid [36]), which may explain this finding. In summary, this study has shown that ET-1 is a potent inhibitor of baseline and stimulated mucin secretion from ovine airway epithelial goblet cells, and that this inhibitory effect is mediated via ETA-receptors. This finding represents the first physiologic mechanism for inhibiting mucin secretion from goblet cells, and, in view of the central role of goblet cells in obstructive pulmonary disease, an understanding of this mechanism may provide novel opportunities for the treatment of these diseases.
The authors are grateful to Tony Chuck, Emma Levett, and Julie Owen for their technical expertise with the radioligand binding and immunohistochemical studies. The advice given by Anthony Davenport was also appreciated. These studies were supported by the Biotechnology and Biological Sciences Research Council and by Pfizer Global Research and Development. Received in original form September 9, 2003 Received in final form August 16, 2004
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
8 cm2). The tissue was pinned, and sterile collagenase A (100 Mandl U/ml) (Roche Diagnostics UK Ltd., Lewes, UK) and dispase 2 (1 Wunsch U/ml) (Roche Diagnostics UK Ltd.) in phosphate-buffered saline (PBS) were injected underneath the epithelium. The preparation was covered in sterile MEM, containing DNase (0.1 mg/ml) (Sigma Chemical Co.) and dithiothreitol (0.15 mg/ml) (Sigma Chemical Co.) (MEM+). After a 30-min incubation at 37°C in a 5% CO2 atmosphere, the epithelium was removed as an intact sheet from the trachea by scraping a cover slip over the surface of the tissue at a 45° angle. The resulting epithelial sheet was suspended in MEM+, cut into suitably sized pieces (5–8 mm diameter) with fine scissors, and mounted, cilia-side up, onto nitrocellulose discs that were coated with collagen (Sigma Chemical Co.). Explants were placed in culture and allowed to attach to the nitrocellulose substrates by overnight incubation (37°C) at the air–liquid interface, with the luminal surface exposed to the humidified 5% CO2 atmosphere. The culture medium used was a 1:1 mixture of Ham's F12 (GIBCO Laboratories) and 3T3-conditioned medium, containing, insulin (10 µg/ml), endothelial growth supplement (7.5 µg/ml), transferrin (5 µg/ml), hydrocortisone (0.036 µg/ml), triiodothyronine (0.02 µg/ml) (all from Sigma Chemical Co.). The resulting explants were utilized the next day for mucin secretion studies. 
