Mucociliary transport in porcine trachea: differential effects of inhibiting chloride and bicarbonate secretion

Abstract

This study was designed to assess the relative importance of Cl⁻ and HCO₃⁻ secretion to mucociliary transport rate (MCT) in ex vivo porcine tracheas. MCT was measured in one group of tissues that was exposed to adventitial HCO₃⁻-free solution while a parallel group was exposed to adventitial HCO₃⁻-replete solution. After measurement of baseline MCT rates, acetylcholine (ACh) was added to stimulate submucosal gland mucous liquid secretion, and MCT rates were again measured. Before ACh addition, the mean MCT was higher in the HCO₃⁻-free group (4.2 ± 0.9 mm/min) than in the HCO₃⁻-replete group (2.3 ± 0.3 mm/min), but this difference was not statistically significant. ACh addition significantly increased MCT in both groups, but ACh-stimulated MCT was significantly lower in the HCO₃⁻-free group (11.0 ± 1.5 mm/min) than in the HCO₃⁻-replete group (17.0 ± 2.0 mm/min). A second series of experiments examined the effect on MCT of blocking Cl⁻ secretion with 100 μM bumetanide. Before adding ACh, MCT in the bumetanide-treated group (1.0 ± 0.2 mm/min) was significantly lower than in the control group (3.8 ± 1.1 mm/min). ACh addition significantly increased MCT in both groups, but there was no significant difference between the bumetanide-treated group (21.4 ± 1.7 mm/min) and control group (19.5 ± 3.4 mm/min). These results indicate that ACh-stimulated MCT has greater dependence on HCO₃⁻ secretion, whereas the basal MCT rate has greater dependence on Cl⁻ secretion.

mucus that is secreted into the lumen of the airways is constantly cleared by the action of cilia located on the apices of airway epithelial cells. This mucociliary transport system serves as a fundamental, innate defense mechanism for the lung. Normally, pathogens and debris that become trapped in airway mucus are rapidly removed from the lung by this system, thereby reducing the need for immune cell recruitment. The importance of mucociliary transport in innate immunity is apparent in diseases where mucociliary transport is compromised. The airways of patients afflicted with primary ciliary dyskinesia, a genetic disease in which deleterious mutations impair ciliary motility, are susceptible to colonization with opportunistic bacteria such as Haemophilus influenzae, Staphylococcus aureus, Streptococcus pneumoniae, and Pseudomonas aeruginosa (22). Persons afflicted with cystic fibrosis (CF) also exhibit reduced mucociliary clearance (32) and increased susceptibility to colonization by a similar spectrum of such pathogens (15). CF is a genetic disease caused by mutations in the gene that codes for the cystic fibrosis transmembrane conductance regulator (CFTR) (33), an anion channel that conducts both Cl⁻ and HCO₃⁻ (27). The underlying cause of the mucociliary transport deficit in CF has been variously attributed to abnormally high rates of transepithelial absorption of Na⁺, which deplete the periciliary liquid (PCL) layer at the airway surface (10), and the loss of an important fraction of transepithelial Cl⁻ and HCO₃⁻ secretion, which is required to secrete liquids onto the airway surfaces and maintain a minimal depth of PCL (5, 21, 25, 39, 43).

Despite the intuitive logic of the airway surface liquid (ASL) “volume hypothesis” for CF pathogenesis, recent evidence suggests that reduced bioavailability of HCO₃⁻ in CF airway liquid may exert deleterious effects unrelated to ASL volume (31). In CF disease, reduced HCO₃⁻ secretion has been demonstrated not only in airway epithelia (8, 38) but also in other exocrine organs, including the human pancreas (12, 14), murine small intestine (7, 37), and murine gallbladder (9); and, the lumina of these organs generally become obstructed with thickened mucus. It was recently hypothesized that loss of functional CFTR in these exocrine tissues reduces HCO₃⁻ concentration in their luminal fluid, thus interfering with the normal unfolding and discharge of mucin macromolecules from mucin-secreting cells and contributing to the formation of the viscous, obstructive mucus characteristic of CF (29). Evidence supporting this notion is drawn from a recent study that demonstrated reduced mucin diffusivity (increased viscosity) at low extracellular HCO₃⁻ concentrations (6).

The present study was designed to assess the relative importance of HCO₃⁻ and Cl⁻ secretion for mucociliary transport. Ex vivo porcine tracheas, which were previously shown to be an effective model for measuring mucociliary transport (5), were used for these experiments. We observed that selective inhibition of HCO₃⁻ and Cl⁻ secretion produced differential effects on mucociliary transport rate that provide new insights into the roles of these secreted anions in mucociliary clearance.

METHODS

Measurement of mucociliary transport.

All procedures with swine were approved by the University of South Alabama Institutional Animal Care and Use Committee and complied with the United States Public Health Service Policy on Humane Care and Use of Laboratory Animals. Young domestic pigs (10–18 kg), obtained from Auburn University Swine Research and Education Center, were sedated by intramuscular injection with ketamine (16 mg/kg) and midazolam (2 mg/kg). Once sedated, the pigs were killed with a lethal overdose of pentobarbital administered through an ear vein. Each trachea was removed intact by sectioning below the larynx and immediately above the first bronchial branch. The trachea was placed in HCO₃⁻-buffered Krebs-Henseleit solution at room temperature, gradually warmed to 37°C over ∼1 h, and then transferred to a separate beaker containing fresh Krebs-Henseleit solution at 37°C for 1 h. The trachea was then removed from its bath, and the trachealis muscle was resected along its length. Each end of the trachea was carefully tied with heavy suture onto two Lucite cannulas that were lightly coated with silicone grease. The cannulas were supported in a rack that held the trachea horizontal in space such that the open slot was oriented uppermost, allowing the ventral mucosal surface to be viewed from above [for a depiction of this apparatus, see online supplement for Ballard et al. (3)]. The rack holding the trachea was placed in a weighted polycarbonate box that was partially submerged in a 37°C heated water bath. The warm (37°C) Krebs-Henseleit incubation solution was added to the box to bathe the outer surface of the trachea, but no solution was allowed to spill into the tracheal lumen. A tempered glass lid was placed on the box to maintain a warm, humidified environment inside the chamber and heated to prevent condensation on the inner surface of the lid. The Krebs-Henseleit bath was constantly bubbled with 5% CO₂ in O₂ gas to maintain solution pH close to 7.4.

After being mounted, the trachea was allowed to stabilize for 30 min. Mucociliary transport was then measured by placing a few fine flakes of dried ink on the caudal mucosal surface of the trachea. A calibrated scale was placed alongside the trachea to track particle movements, which were captured with an analog video camera and recorded on video cassettes for offline analysis. Mucociliary transport was measured for three consecutive 20-min “runs.” For each run, the transport rate was determined for the first visible moving particle. At the end of the third run, mucus, gently collected from the cranial end of the trachea with Eppendorf pipettes, and bath samples were taken for HCO₃⁻ analysis. Next, 100 μM acetylcholine (ACh) was added to the bath to induce submucosal gland mucus secretion, and 20 min were allowed for the secretory response to fully develop. Mucociliary transport rates for three additional 20-min runs were determined. At the end of the third run, mucus and bath samples were again taken for HCO₃⁻ analysis. The particles usually began to move at different times. Particles that were even within a few hundred microns of each other often began to move at different times; but, once moving, most particles appeared to move at approximately the same rate. Some particles did not move at all. In our previous experience with this technique, we noted that the number of static particles appeared to depend on the treatment (agents that reduce liquid secretion rate tend to increase the number of static particles, whereas inducers of liquid secretion rate tend to reduce the number of static particles) or the size (larger particles or aggregates of particles appear to be more prone to stasis). The rate of the first particle to move out ahead of the other particles was chosen as the rate value for that time period. We reasoned that there is much less chance that a particle with a fast rate would be the consequence of artifact, whereas a static or slower-moving particle could simply be stuck on the surface or its motion inhibited by undetected surface features of the mucosal surface.

To inhibit HCO₃⁻ secretion, HCO₃⁻-free Krebs-Henseleit solution was substituted for the HCO₃⁻-replete Krebs-Henseleit bath solution throughout the experiments. In these experiments, the HCO₃⁻-free Krebs-Henseleit solution was bubbled with 100% O₂ gas. To inhibit Cl⁻ secretion, 100 μM bumetanide, a selective blocker of the basolateral membrane Na⁺-K⁺-2Cl⁻ cotransporter (NKCC) that mediates the active step in transepithelial Cl⁻ secretion in airway epithelia, was added to the HCO₃⁻-replete Krebs-Henseleit bath solution at the beginning of the 1-h incubation period before cannulation and was present throughout the remainder of the experiment. An equal volume of the bumetanide vehicle, dimethyl sulfoxide, was added to the bath solution of control experiments. When possible, control experiments for both HCO₃⁻-free and bumetanide treatments were performed on alternate days to minimize possible temporal variations in mucociliary transport between the preparations.

To evaluate the effect of HCO₃⁻ removal on the buffering capacity of the bath solution, time-control experiments were performed with two tracheas. One trachea was bathed in normal HCO₃⁻-replete Krebs-Henseleit solution, and the other trachea was bathed in HCO₃⁻-free Krebs-Henseleit solution. Time-control experiments included addition of ACh. Bath solution pH was monitored at regular intervals with a pH electrode (SevenEasy pH meter; Mettler-Toledo) throughout the time frame of a typical experiment.

Measurement of bath HCO₃⁻ concentrations.

Mucus and bath samples were collected, placed in tubes, immediately sealed, and refrigerated at 4°C until analyzed. HCO₃⁻ concentrations were measured using the Infinity CO₂ kit (Thermo Scientific, West Sussex, UK) and a Heλios α-spectrophotometer (Thermo Electron).

Solutions.

Krebs-Henseleit solution was composed of 125.0 mM NaCl, 25.0 mM NaHCO₃, 1.0 mM CaCl₂·2H₂0, 1.0 mM MgSO₄, 2.12 mM K₂HPO₄, 0.375 mM KH₂PO₄, and 10 mM glucose. This solution was gassed with 5% CO₂ in O₂. To make HCO₃⁻-free Krebs-Henseleit solution, NaHCO₃ was omitted, and NaCl concentration was increased to 150 mM so that the two solutions were approximately isosmotic. Aqueous NaOH was used to titrate the pH of HCO₃⁻-free Krebs-Henseleit solution to 7.4. HCO₃⁻-free Krebs-Henseleit solution was gassed with 100% O₂.

Statistics.

Data are expressed as means ± SE. “n” refers to the number of tracheas (i.e., animals) from which measurements were made. SigmaPlot 12 software (Sigmastat Software, San Jose, CA) was used for statistical comparisons. The threshold for significance was P < 0.05. When appropriate, comparisons were made with paired t-tests; otherwise, unpaired t-tests were used. If the test for normality failed, the Wilcoxon sign rank test was used for comparisons.

RESULTS

Effect of HCO₃⁻ removal on mucociliary transport.

In Fig. 1A, the effect of HCO₃⁻ removal from the bath solution on mucociliary transport rate was measured. Tracheas were bathed in either normal HCO₃⁻-replete or HCO₃⁻-free Krebs-Henseleit buffer. During the pre-ACh period, there was no significant difference in mucociliary transport rates between the HCO₃⁻-replete and the HCO₃⁻-free groups at any of the three 20-min time periods. Following addition of ACh, mucociliary transport rates were substantially increased for both groups. There was no significant difference in mucociliary transport rates between the two groups at the first post-ACh period; but, in the final two periods, the mucociliary transport rates were significantly lower in the HCO₃⁻-free group. Figure 1B shows the absolute change in mucociliary transport from the periods immediately before and after ACh addition. The mucociliary transport rate response in tracheas exposed to HCO₃⁻-free solution was significantly lower than in tissues exposed to HCO₃⁻-replete solution. Figure 1C shows mucociliary transport rates averaged from the three pre-ACh periods (60 min) and three post-ACh periods (60 min) for tracheas exposed to HCO₃⁻-replete or HCO₃⁻-free solutions. Addition of ACh significantly increased the mucociliary transport rate in both HCO₃⁻-replete and HCO₃⁻-free solutions. The post-ACh response to HCO₃⁻-free solution was significantly less than the response to HCO₃⁻-replete solution. Although the mean pre-ACh mucociliary transport rates appeared somewhat higher in HCO₃⁻-free solution than in HCO₃⁻-replete solution, this difference was not statistically significant. These responses indicate that loss of HCO₃⁻ secretion has no significant effect on the basal rate of mucociliary transport but significantly reduces the ACh-induced stimulation of mucociliary transport.

Fig. 1.

Effects of HCO₃⁻-free solution on tracheal mucociliary transport rates. Mucociliary transport rates were measured for each of three consecutive 20-min periods. Next, 100 μM acetylcholine (ACh) was applied, and mucociliary transport rates were measured for an additional three consecutive 20-min periods. A: experiments performed with tracheas exposed to HCO₃⁻-replete Krebs-Henseleit solution (n = 10) or HCO₃⁻-free Krebs-Henseleit solution (n = 10). *Transport rates were significantly (P < 0.05) lower than rates at the same time points for the HCO₃⁻-replete group. B: absolute changes in mucociliary transport rates following ACh addition. Bars represent the differences in mucociliary transport rates between the periods immediately before and after ACh addition for both HCO₃⁻-replete and HCO₃⁻-free treatment groups. *The response to ACh was significantly (P < 0.05) lower in the HCO₃⁻-free group. C: comparison of aggregate mucociliary transport rates before and after ACh addition. Bars represent the means of mucociliary transport rates for the three periods before and the three periods after ACh addition for both HCO₃⁻-replete and HCO₃⁻-free treatment groups. *Mucociliary transport rates were significantly (P < 0.05) increased following ACh application for both HCO₃⁻-replete and HCO₃⁻-free treatment groups. Cross indicates that the response to ACh addition was significantly (P < 0.05) lower in the HCO₃⁻-free group than the HCO₃⁻-replete group.

To assess the acid-buffering effects of HCO₃⁻ removal from the Krebs-Henseleit solution, time-control experiments were performed where pH was monitored at regular intervals throughout the protocol. After tracheas were initially warmed to 37°C, their placement in either bath solution resulted in a transient acidification of the bath followed by a gradual recovery. Recovery of pH took longer for the HCO₃⁻-free Krebs-Henseleit than for the HCO₃⁻-replete Krebs-Henseleit solution. For the periods during which mucociliary transport was measured, the pH in the HCO₃⁻-replete solution ranged from 7.46 to 7.47 and the HCO₃⁻-free solution from 7.26 to 7.36.

Effect of inhibiting Cl⁻ secretion on mucociliary transport.

To inhibit transepithelial secretion of Cl⁻, tracheas were pretreated with 100 μM bumetanide and bathed in HCO₃⁻-replete Krebs-Henseleit solution. The basal rate of mucociliary transport before addition of ACh to the bumetanide-pretreated tracheas was very low (Fig. 2A). Although the rates during the first period were not significantly different, the basal rates in the second and third periods of the bumetanide-treated group were significantly lower than the control group. There was no significant difference in ACh-stimulated mucociliary transport rates between the bumetanide-pretreated and control tracheas during any of the stimulated periods. Figure 2B shows that there was no significant difference in the absolute change in the mucociliary transport rate between the bumetanide-treated and untreated tracheas for the periods immediately before and after stimulation. Figure 2C shows the mean mucociliary transport rates averaged for the three pre-ACh periods (60 min) and three post-ACh periods (60 min) for tracheas in the presence or absence of bumetanide. The mucociliary transport rates were significantly lower in the bumetanide-treated tracheas than the controls before ACh addition. However, there was no significant difference between the bumetanide and control groups after ACh addition. These results indicate that inhibition of Cl⁻ secretion with bumetanide significantly reduces the basal mucociliary transport rate but has no effect on the ACh-stimulated mucociliary transport rate.

Fig. 2.

Effects of bumetanide on trachea mucociliary transport rates. All tissues were bathed in HCO₃⁻-replete Krebs-Henseleit bath solution. Mucociliary transport rates were measured for each of three consecutive 20-min periods. Next, 100 μM ACh was applied, and mucociliary transport rates were measured for an additional three consecutive 20-min periods. A: tracheas were pretreated with either 100 μM bumetanide (n = 8) or the vehicle (n = 6). *Significant (P < 0.05) difference between bumetanide-treated tracheas and control tracheas for the same time points. B: absolute changes in mucociliary transport rates following ACh addition. Bars represent differences in mucociliary transport rates between the periods immediately before and after ACh addition for both bumetanide and control pretreatment groups. No significant difference was observed between the two groups. C: comparison of aggregate mucociliary transport rates before and after ACh addition. Bars show the sums of mucociliary transport for the three periods before and the three periods after ACh addition for both bumetanide and control treatment groups. *Mucociliary transport rates were significantly (P < 0.05) increased following ACh application for both bumetanide and control pretreatment. Cross indicates that bumetanide pretreatment significantly (P < 0.05) reduced the mucociliary transport rate before ACh addition.

Mucus HCO₃⁻ concentrations.

As expected, the HCO₃⁻ concentrations in the luminal mucus were substantially and significantly lower in the tracheas exposed to nominally HCO₃⁻-free solution than mucus from tracheas bathed in the HCO₃⁻-replete solutions, both before and after ACh stimulation (Fig. 3A). In addition, the HCO₃⁻ concentration in the luminal mucus was significantly higher than the bath solution in 1) the pre-ACh HCO₃⁻-replete group (mucus: 25.7 ± 1.8 meq/l, n = 10; bath: 21.1 ± 0.5 meq/l, n = 10); 2) the post-ACh HCO₃⁻-replete group (mucus: 28.1 ± 1.5 meq/l, n = 10; bath: 21.1 ± 0.4 meq/l, n = 10); and 3) the pre-ACh HCO₃⁻-free group (mucus: 3.4 ± 0.8 meq/l, n = 9; bath: 0.1 ± 0.1 meq/l, n = 10). The HCO₃⁻ concentration in the mucus of the post-ACh HCO₃⁻-free group was not significantly different from the concentration in the bath (mucus: 1.4 ± 1.4 meq/l, n = 10; bath: 0.3 ± 0.1 meq/l, n = 10).

Fig. 3.

Luminal mucus HCO₃⁻ concentrations. Mucus was collected from the mucosal lumen at the cranial end of the trachea following the third pre-ACh and third post-ACh periods. A: effect of HCO₃⁻-free bath solution on mucus HCO₃⁻ concentration. *Significant (P < 0.05) reduction in HCO₃⁻ concentration in the tracheas exposed to HCO₃⁻-free bath solution compared with mucus from tracheas exposed to HCO₃⁻-replete bath solution. B: effect of bumetanide pretreatment on mucus HCO₃⁻ concentration. **Significant (P < 0.05) increase in mucus HCO₃⁻ concentration in the post-ACh bumetanide group compared with both the post-ACh control group and the pre-ACh bumetanide.

Mucus HCO₃⁻ concentrations did not significantly change following ACh addition to control tracheas (Fig. 3B). However, the mucus HCO₃⁻ concentrations in the bumetanide-treated tracheas significantly increased 2.3-fold following ACh addition. Mucus HCO₃⁻ concentrations were not significantly different from the bath solutions in 1) pre-ACh control tracheas (mucus: 22.4 ± 2.5 meq/l, n = 6; bath: 21.9 ± 1.9 meq/l, n = 6); 2) pre-ACh bumetanide-treated tracheas (mucus: 27.0 ± 3.1 meq/l, n = 8; bath: 22.8 ± 1.4 meq/l, n = 8); and 3) post-ACh control tracheas (mucus: 31.3 ± 5.3 meq/l, n = 6; bath: 20.4 ± 1.1 meq/l, n = 6). However, mucus HCO₃⁻ concentrations were significantly greater than the bath solution in the post-ACh bumetanide-treated tracheas (mucus: 62.0 ± 6.4 meq/l, n = 8; bath: 22.2 ± 0.9 meq/l, n = 8).

DISCUSSION

The objective of this study was to assess the individual impacts of HCO₃⁻ and Cl⁻ secretion on mucociliary transport in pulmonary airways. Our results demonstrate that blockade of HCO₃⁻ secretion significantly reduces the ACh-stimulated mucociliary transport rate but has no significant effect on the basal mucociliary transport rate. We also find that the basal rate of mucociliary transport is significantly reduced when Cl⁻ secretion is inhibited by the loop diuretic bumetanide, but this inhibitor has no apparent effect on the ACh-stimulated mucociliary transport rate. These results suggest that the stimulatory effects of ACh on mucociliary transport are linked to the secretion of HCO₃⁻ from submucosal glands and that compromise of this component fractionally impairs this important innate defense response. Furthermore, our findings indicate that mucociliary transport must be largely dependent upon bumetanide-sensitive Cl⁻ secretion in the absence of submucosal gland stimulation. Because ACh induces mucous liquid secretion predominantly if not entirely from submucosal glands (4, 28), these data suggest that basal mucociliary transport activity is largely dependent upon Cl⁻-driven liquid secretion, perhaps from the surface epithelium, whereas the increased mucociliary transport seen with ACh is at least partially supported by HCO₃⁻-dependent liquid secretion that originates from submucosal glands. This notion is supported by the observations of Inglis et al. (18), who reported that ACh induces an alkaline burst by isolated perfused porcine bronchi.

There are at least three factors that potentially account for the pattern of inhibition of mucociliary transport that we observed with bumetanide and HCO₃⁻-free solutions. First, our manipulations of ion transport could have directly reduced the beat frequency of cilia that propel airway mucus toward the larynx. Second, the responses could have been due to inhibition of anion secretion and the resulting decrease in ASL volume and depth. Third, removal of HCO₃⁻ from the extracellular solutions could have interfered with the normal process of mucin secretion, thus altering the physical properties of mucus as recently proposed (31). Each of these possibilities is discussed below.

Transport of mucus over airway surfaces toward the larynx is accomplished by coordinated beating of cilia. Ciliary beat frequency (CBF) in mammalian airways is variable, but typically ranges between 5 and 15 Hz (23, 34, 46, 47). When muscarinic agonists are applied to airway epithelia, CBF increases approximately twofold (46, 47); however, this effect is transient, with CBF returning to near baseline levels within 20–30 min (46). Furosemide, which like bumetanide inhibits NKCC and transepithelial Cl⁻ secretion, has no appreciable effect on tracheal CBF when administered in vivo to dogs or baboons (45) or in vitro to rabbits (40). Consequently, it seems unlikely that the effects on mucociliary transport observed in the present study are due to changes in CBF caused by ACh or bumetanide.

On the other hand, ciliated epithelial cells express an HCO₃⁻-sensitive, soluble adenylyl cyclase in the apical region that increases ciliary activity through production of cAMP when cytoplasmic HCO₃⁻ concentrations increase (35, 36). Thus, extracellular HCO₃⁻-free solution may have reduced both the intracellular HCO₃⁻ concentration and CBF via reduced kinase activity. However, we observed that the mean basal mucociliary transport rate in pig tracheas exposed to HCO₃⁻-free solution was unchanged or actually increased, albeit insignificantly, compared with controls during the period before ACh stimulation (Fig. 1). Furthermore, we observed that mucociliary transport increased in the absence of HCO₃⁻ when stimulated with ACh, albeit significantly less than in the presence of HCO₃⁻, indicating that CBF unlikely decreased even without HCO₃⁻ in the medium. Nonetheless, we cannot discount the possibility that HCO₃⁻ removal fractionally inhibited pig tracheal CBF through this cAMP-sensitive pathway but that the effect was too small to distinguish in the absence of ACh. It is worth noting that airway cilia are relatively insensitive to extracellular pH changes within physiological limits. In human bronchial explants bathed in HCO₃⁻-free, phosphate-buffered saline, CBF was found to be within 10% of maximum (∼13 Hz) between extracellular pH of 6.5 and 9.5 (23).

Several studies have documented that mucociliary transport rate varies with the rate of airway liquid secretion. In studies with cultured monolayers of human ciliated bronchial epithelial cells, Tarran et al. (41) demonstrated that mucociliary transport rates correlated with ASL depth, i.e., reduced ASL depth was associated with reduced mucociliary transport rates. In intact bronchial airways, muscarinic agonists induce a voluminous secretion of liquid that largely, if not entirely, originates from submucosal glands (4). These agonists drive liquid secretion through induction of both Cl⁻ and HCO₃⁻ secretion (2). Secretion of both anions depends upon the presence of anion channels, likely both CFTR and TMEM16A, in the apical membrane of gland serous cells (21). Transepithelial Cl⁻ secretion requires entry of Cl⁻ across the basolateral membrane through bumetanide-sensitive NKCC and exit through apical membrane anion channels (2). HCO₃⁻ and H⁺ are generated intracellularly by carbonic anhydrase from CO₂ and H₂O. Extrusion of H⁺ across the basolateral membrane is required for the net intracellular production of HCO₃⁻; thus, HCO₃⁻ secretion can be effectively blocked by drugs that inhibit Na⁺/H⁺ exchangers such as dimethylamiloride (2). HCO₃⁻ also exits the cells across the apical membrane through the anion channels (21). Cholinergic stimulation increases the total depth of ASL about threefold in ex vivo bovine tracheas (48).

In porcine bronchi, bumetanide-inhibitable Cl⁻ secretion accounts for 70% of ACh-induced mucus liquid secretion, whereas dimethylamiloride-sensitive HCO₃⁻ secretion is responsible for only 9% of this total (42). The combination of these two inhibitors blocks 89% of the total secretory response to ACh (42) and profoundly inhibits mucociliary transport in tracheas even in the presence of ACh (5), demonstrating the dependence of mucociliary transport rate on airway liquid secretion. In the present study, bumetanide pretreatment significantly reduced mucociliary transport, which is consistent with this inhibitor's blockade of a substantial fraction of the basal liquid secretion. However, ACh induced a substantial significant increase in mucociliary transport in the presence of bumetanide, which should have inhibited a large fraction of the liquid secretion. We propose that this paradox can be explained by the following. The rates of liquid secretion that we previously documented in cannulated bronchi were 1.9 ± 0.3 μl·cm⁻²·h⁻¹ for unstimulated tissues and 16.9 ± 2.7 μl·cm⁻²·h⁻¹ for ACh-treated tissues (42). If bumetanide blocks 70% of the liquid secretion responses in both cases, then the tissues without ACh would secrete 0.6 μl·cm⁻²·h⁻¹ in the presence of bumetanide, and the ACh-treated tissues would secrete 5.1 μl·cm⁻²·h⁻¹ in the presence of bumetanide. It is reasonable to expect that the lower liquid secretion rate with bumetanide pretreatment before ACh would reduce mucociliary transport. However, in the presence of ACh, the bumetanide-pretreated tissues should still secrete about 2.7 times more liquid than in the absence of ACh. This rate may well be capable of stimulating mucociliary transport. Data collected from a previous study of ACh-stimulated bronchi (24) indicated that bumetanide pretreatment increased the percent mucous solids only by about 36% above that collected from the control airways. We doubt that this modest increase in percent solids with bumetanide is sufficient to substantially affect mucociliary transport rate.

Last, the possible effect of HCO₃⁻ removal on the physical properties of airway mucus must be considered as a potential cause of its reduced transportability. Acidification of mucus is reported to substantially increase its viscosity (16, 17), which could reduce ciliary clearance (20). A recent study demonstrated that the viscosity of secreted mucins is inversely related to the HCO₃⁻ concentration in the medium and that expansion of mucins upon release from secretory granules is a function of extracellular HCO₃⁻ (6). High concentrations of Ca²⁺ and H⁺ are thought to be required to stabilize condensed polyelectrolyte mucin polymers within secretory granules by electrostatically shielding and cross-linking fixed anionic residues on mucins (44) and by coordinated binding of Ca²⁺-specific protein and/or interactive hydrophobic bonds in binding nodes (1, 19). Because HCO₃⁻ sequesters Ca²⁺ (6) and may competitively displace Ca²⁺ from the electrostatic bridges and nodal binding sites in condensed mucin, this anion may facilitate the normal unfolding and maturation of the mucin molecules at the moment of granule rupture (30, 31). By extrapolation, reduction in HCO₃⁻ concentration in extracellular liquid experimentally (by inhibition of HCO₃⁻ secretion) or pathologically (as might occur from the loss of CFTR channel function in CF) could result in incomplete mucin expansion upon exocytosis, resulting in a more dense, viscous mucus. Not only does reduced HCO₃⁻ restrict mucin expansion (6), but recent findings by Gustafsson et al. (13) demonstrate that mucus viscosity appears to be abnormally high in CF ileal mucus and that supplemental HCO₃⁻ appears to restore the mucus properties toward normal. If insights may be taken from CF, where both mucus and HCO₃⁻ secretion are abnormal not only in the airways but also in mucus-secreting organs in general (31), it is tempting to conclude that the lack of HCO₃⁻ contributes significantly to a change in mucus physical properties that impede its transport (26).

Our present study demonstrates that removing HCO₃⁻ from the bath solution impairs the rate of mucociliary transport only when submucosal glands were stimulated with ACh. We also observed that pretreatment of tracheas with bumetanide, an inhibitor of Cl⁻ secretion, had no effect on ACh-stimulated mucociliary transport, which seemingly runs counter to the notion that ASL volume dominates control of mucociliary clearance. As previously reported (42), we observed that bumetanide pretreatment substantially elevated mucus liquid HCO₃⁻ concentrations. These elevated HCO₃⁻ concentrations could have ameliorated the inhibitory actions of volume reduction on mucociliary transport by potentially improving mucin viscosity and transportability despite decreases in ASL volume. Resolution of this important issue warrants further study.

In our experiments, Cl⁻ was substituted for HCO₃⁻ to minimally alter the ionic composition of our solutions. However, this maneuver fractionally reduced the buffering capacity of the HCO₃⁻-free Krebs-Henseleit solutions. When the pHs of the HCO₃⁻-free and HCO₃⁻-replete Krebs-Henseleit solutions were monitored in time-control experiments, the bath transiently acidified following immersion of the intact tracheas into the bath. The pH in both solutions gradually recovered over time; however, the HCO₃⁻-free solution recovered more slowly. For the periods during which mucociliary transport was measured, the pH in the HCO₃⁻-replete solution ranged from 7.46 to 7.47 and the HCO₃⁻-free solution from 7.26 to 7.36. Although the bath pHs in the HCO₃⁻-free solution were lower, we expect that these pHs are probably within acceptable limits since pHs in this range 1) rarely cause symptoms in humans apart from chronic decalcification of bones (11), 2) should not affect the fluid secretion response to ACh (Ballard, unpublished observations), and 3) should not affect CBF (23). It should be noted that, at the time points where mucociliary transport rates would have been measured in the presence of ACh [the only time points that were significantly different from the HCO₃⁻-replete solution (Fig. 1)], the pHs in the HCO₃⁻-free solution were 7.35–7.37. Consequently, we believe that this lower buffering capacity of the HCO₃⁻-free solution minimally affected our results.

In summary, we report the relative effects of independently inhibiting Cl⁻ and HCO₃⁻ secretion on mucociliary transport. We found that inhibition of Cl⁻ secretion with bumetanide decreased basal mucociliary transport rates but had no effect on ACh-stimulated rates, whereas removing HCO₃⁻ had no significant effect on basal rates but significantly depressed the ACh-stimulated mucociliary transport rates. We speculate that HCO₃⁻-dependent changes in mucus physical properties most likely account for these observations.

GRANTS

This work was funded by the Cystic Fibrosis Foundation Therapeutics grant BALLAR07XX0. S. T. Ballard also received support from National Heart, Lung, and Blood Institute (NHLBI) Grant R01HL-063302. P. M. Quinton is funded by Cystic Fibrosis Foundation Therapeutics Grant QUINTO08XX0, NHLBI Grant R01HL-84042, and the Nancy Olmsted Endowment.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: J.L.C. performed experiments; J.L.C. and S.T.B. analyzed data; J.L.C., P.M.Q., and S.T.B. edited and revised manuscript; J.L.C., P.M.Q., and S.T.B. approved final version of manuscript; P.M.Q. and S.T.B. conception and design of research; P.M.Q. and S.T.B. interpreted results of experiments; S.T.B. prepared figures; S.T.B. drafted manuscript.