Mechanisms of airway responses to esophageal acidification in cats

We describe the neurochemical mechanisms for activation of the airway responses to acid exposure of the esophagus only. The airway constrictive response, which does not produce a change in airway resistance, is mediated by vagal cholinergics. The increased mucous secretion is mediated by a local neural pathway that primarily affects the airway apocrine glands. The low amplitude of these responses suggests that they are part of the innate, rather than acute emergency, defense system.

Abstract

Acid in the esophagus causes airway constriction, tracheobronchial mucous secretion, and a decrease in tracheal mucociliary transport rate. This study was designed to investigate the neuropharmacological mechanisms controlling these responses. In chloralose-anesthetized cats (n = 72), we investigated the effects of vagotomy or atropine (100 μg·kg⁻¹·30 min⁻¹ iv) on airway responses to esophageal infusion of 0.1 M PBS or 0.1 N HCl at 1 ml/min. We quantified 1) diameter of the bronchi, 2) tracheobronchial mucociliary transport rate, 3) tracheobronchial mucous secretion, and 4) mucous content of the tracheal epithelium and submucosa. We found that vagotomy or atropine blocked the airway constriction response but only atropine blocked the increase in mucous output and decrease in mucociliary transport rate caused by esophageal acidification. The mucous cells of the mucosa produced more Alcian blue- than periodic acid-Schiff (PAS)-stained mucosubstances, and the mucous cells of the submucosa produced more PAS- than Alcian blue-stained mucosubstances. Selective perfusion of the different segments of esophagus with HCl or PBS resulted in significantly greater production of PAS-stained mucus in the submucosa of the trachea adjacent to the HCl-perfused esophagus than in that adjacent to the PBS-perfused esophagus. In conclusion, airway constriction caused by esophageal acidification is mediated by a vagal cholinergic pathway, and the tracheobronchial transport response is mediated by cholinergic receptors. Acid perfusion of the esophagus selectively increases production of neutral mucosubstances of the apocrine glands by a local mechanism. We hypothesize that the airway responses to esophageal acid exposure are part of the innate, rather than acute emergency, airway defense system.

NEW & NOTEWORTHY

We describe the neurochemical mechanisms for activation of the airway responses to acid exposure of the esophagus only. The airway constrictive response, which does not produce a change in airway resistance, is mediated by vagal cholinergics. The increased mucous secretion is mediated by a local neural pathway that primarily affects the airway apocrine glands. The low amplitude of these responses suggests that they are part of the innate, rather than acute emergency, defense system.

prior studies have found that gastroesophageal reflux is sometimes associated with symptoms of asthma (13, 23), but whether these symptoms were caused by stimulation of the esophagus or supraesophageal structures was unknown. No study of this issue (28, 34, 58) had prevented supraesophageal reflux, and acid exposure of the larynx (28) can cause significant changes in airway function. We were the first to examine the respiratory effects of esophageal acid exposure while preventing supraesophageal reflux of acid (31). We found that esophageal acid exposure significantly decreased the diameter of the smaller airways without causing a change in airway resistance, suggesting that this reflex response does not cause symptoms of asthma. Therefore, we concluded that the airway changes reflexly activated by esophageal acid exposure likely serve a physiological, rather than an emergency, protective function or cause a pathological condition (31).

Our other findings suggested the possible physiological function of the mild bronchoconstriction caused by the esophageal reflex activated by acid. We found that exposure of the esophagus to acid caused a decrease in tracheal mucociliary transport rate and an increase in tracheobronchial mucous output. Tracheal mucociliary transport rate is indirectly related to the amount of epithelial mucus secreted (43, 49); therefore, it is likely that the decrease in the rate of mucociliary transport caused by esophageal acidification was due in part to the increase in mucous output. Considering that mucus is an excellent buffer of acid (15), we hypothesized that an acid-induced esophageal reflex caused mild constriction of small airways, without increasing airway resistance, which assists the expulsion of mucus from the airway mucous glands in preparation for microaspiration of refluxed acid.

It has been reported that cutting of the vagus nerves blocked the fall in airway conductance caused by esophageal acidification (34). Dogs with esophagitis were used in these studies, and no effort was made to prevent supraesophageal reflux of the acid infused into the esophagus. In these studies (34), one cannot exclude the possibility that the airway response to the infusion of acid was due to supraesophageal reflux of acid, rather than activation of an esophageal reflex. Therefore, the role of the vagus nerves in the reflex activation of the airway responses to stimulation of the esophagus with acid is unknown.

The role of cholinergic receptors in the airway motor responses to esophageal acidification is unclear. In humans (24), but not rabbits (19), atropine inhibited the increase in airway resistance caused by esophageal acidification. However, these studies in humans (24) cannot be considered specific to the esophagus, because no attempt was made to prevent supraesophageal reflux of acid, and laryngeal acidification-induced airway constriction is mediated by the vagus nerves (58). In rabbits (19), very high concentrations of HCl (up to 4 N) were used, such that the acid may have had a direct and/or nonphysiological effect on the airways. Therefore, the role of acetylcholine receptors in the airway responses to esophageal acidification is unknown.

Airway mucous secretion is under vagal control (16, 18, 21, 50), but the mechanism of increased airway mucous secretion in response to reflex stimulation of the esophagus by acid is unknown. Cholinergic receptors have been found to mediate airway mucous secretion stimulated reflexly (21, 50) or by vagal stimulation (16, 18, 47), but the neuropharmacological receptors involved in the airway mucous response to reflex stimulation of the esophagus by acid are unknown.

The airway is not only under extrinsic long arc neural control, but it is also under extrinsic direct neural control from the esophagus. Two types of direct neural pathways from the esophagus to the airways have been identified: one pathway, which causes a decrease in tone (6, 14), is connected with airway smooth muscle, and a separate pathway is connected with the tracheal ganglia (35). Considering that the tracheal ganglia supply innervation to all effector cells, this esophagotracheal connection could mediate glandular, i.e., mucous, as well as muscular, responses. The role of this direct neural pathway in the mucous response to esophageal acidification has not been investigated.

There are two sources and histochemical types of mucus in the airway (17, 18, 60): acid mucins of the mucosal goblet cells and neutral mucins of the submucosal apocrine glands. However, the source and type of mucus produced and secreted by the airway in response to esophageal acidification are unknown.

The aims of this study were to investigate whether the vagus nerves, the muscarinic cholinergic receptors, or the local connections between the esophagus and trachea mediate any of the airway responses to esophageal acidification and to determine the source of the mucus secreted in response to esophageal acidification. We hypothesize that vagus nerves, local connections between the esophagus and trachea, and the muscarinic receptors play roles in the airway responses to acid exposure of the esophagus. In addition, we hypothesize that the submucosal apocrine glands represent the primary source of secreted mucins in response to esophageal acid exposure.

MATERIALS AND METHODS

Animal Preparation

General preparation.

Cats (n = 72, 2.2-4.7 kg body wt) of either sex were anesthetized using α-chloralose. α-Chloralose was chosen because we have found in >20 years of research (37) that this is the best anesthetic available to provide a surgical level of anesthesia while preserving reflex responsiveness in cats. α-Chloralose was used in a specific manner and in conjunction with other anesthetics, as described below, to obtain a surgical level of anesthesia in all cases, and different protocols were used based on the extent of surgical preparation. The specific anesthetic protocols were selected in consultation with and were approved by the Institutional Animal Care and Use Committee of the Medical College of Wisconsin.

For the airway diameter, mucociliary transport, and mucous secretion studies, α-chloralose was titrated to effect to ensure that all animals reached the same surgical level of anesthesia. α-Chloralose was initially given at 55 mg/kg ip, and if the animals had not attained a surgical level of anesthesia in 1 h, they were supplemented with 25% of the original dose. If after 45 min they were still not at a surgical level of anesthesia, pentobarbital sodium (3 mg/kg iv) was administered. Less than 5% of the animals required pentobarbital sodium. For the experiments involving histology, the cats were placed in an anesthetizing box and anesthetized using isoflurane (5%). Once anesthetized, the cats were removed and placed on their back, and anesthesia was maintained with isoflurane (3%) via a mask. After the appropriate surgeries were performed (see below), α-chloralose (60 mg/kg iv) was injected slowly as the concentration of isoflurane was reduced to maintain anesthesia during the experiments. Experiments were begun ≥1 h after isoflurane was discontinued.

The femoral vein was cannulated for infusion of 0.9% NaCl for hydration. In all cats, 0.5-cm-diameter Tygon tubing was used to cannulate the esophagus just distal to the upper esophageal sphincter and just proximal to the lower esophageal sphincter. The opposite end of the upper esophageal sphincter cannula exited the oral cavity. The opposite end of the lower esophageal sphincter cannula exited the digestive tract through the stomach wall and the abdomen through the abdominal incision. A gastric fistula was created in the fundus to allow continuous evacuation of gastric secretions. The animals were killed after the study using a pentobarbital sodium-based euthanasia solution (600 mg iv).

Specific preparations.

VAGOTOMY.

The vagus nerves were carefully separated from the carotid sheaths, and a thread was placed, but not tied, around the vagus nerves. These threads exited through the neck wound and were used during the study to pull the vagi out of the neck when appropriate to allow sectioning of the vagi without disturbing the trachea or esophagus.

BRONCHIAL DIAMETER.

Tantalum bronchograms were obtained as described previously (31). The trachea was intubated, and the cats were placed supine onto the microfocus X-ray system stage for imaging of the bronchi. The bronchi were dusted with fine tantalum powder injected into the air inflow from a ventilator set at two to three times tidal volume for three to five breaths. The goal was to finely and evenly coat the airway with tantalum dust. After the airway was dusted, it was imaged, and the study was begun. Imaging data were saved on computer for postacquisition analysis.

Tantalum has been found to be chemically inert (39), causing no inflammation of the airways or lungs and no significant changes in blood gases or measures of lung function; in our prior study (31) we found that lung mechanics did not change in response to esophageal acidification whether or not the airways or lungs were exposed to tantalum.

TRACHEOBRONCHIAL MUCOCILIARY TRANSPORT.

Tracheobronchial mucociliary transport was measured as described previously (31). Cats were placed onto the microfocus X-ray imaging system stage, and a small hole was made in a tracheal cartilage ring using a cautery. Care was taken to avoid bleeding into the trachea, as bleeding retarded mucociliary transport. A tantalum disk injection device was inserted into the hole in the tracheal cartilage, and tantalum disks (1 mm diameter, 0.3 μg) were injected as far distally into the airway as possible. The disks usually settled near the carina. Disk movement through the trachea by mucociliary transport was recorded using the microfocus X-ray system, and the data were stored on computer. Later, the velocity of orad travel of the tantalum disks was measured. Between injections of disks, a saline-soaked gauze pad was placed over the hole in the tracheal cartilage. After transiting through the trachea, the tantalum disks settled on the hard palate. An endotracheal tube was not placed in these studies in order to disturb the trachea as little as possible.

TRACHEAL MUCOUS OUTPUT.

The trachea was transected just distal to the cricoid cartilage, and a Y-shaped endotracheal tube was inserted. The cat was placed at a 10° angle head-down to promote outflow by gravity. Polyethylene (PE-50) tubing attached to a syringe pump was inserted into the nondependent arm of the endotracheal tube until resistance was met. The catheter was then withdrawn 0.5 cm and taped to the endotracheal tube. On average, the tube was inserted ∼10 cm from the larynx. Saline (0.9% NaCl) was injected at 0.25 ml/min, and the perfusate was collected every 30 min over ice. Volumes of the samples were determined, and hexosamine concentrations were quantified.

EFFECT OF SELECTIVE PERFUSION OF THE ESOPHAGUS.

The esophagus was ligated 4 cm from the cricoid cartilage, forming two esophageal sections, proximal and distal. Tubing (PE-50) was inserted into the ends of the larger catheters already implanted (as described above) and advanced to the point of the esophageal ligation. This arrangement allowed for selective perfusion of each esophageal segment as fluid entered the esophageal sections through the smaller catheters and exited through the larger catheters.

Protocols

In all studies the esophagus was perfused with 0.1 M PBS (PBS) or 0.1 N HCl (HCl) at 1 ml/min. When atropine was given, it was administered at 100 μg/kg iv every 30 min until the end of the experiment.

Role of vagus nerves or cholinergic receptors.

BRONCHIAL DIAMETER.

The esophagus was perfused with PBS or HCl throughout the experiment, and microfocus X-ray images of the bronchus were taken every 30 min to quantify bronchial diameter. The esophagus was perfused for 1 h with PBS, and the vagus nerves were transected or atropine was administered after the first 30 min of PBS perfusion. The perfusate was changed to HCl for 30 min and then to PBS for 30 min.

TRACHEAL MUCOCILIARY TRANSPORT.

While microfocus X-rays of the trachea were obtained, the esophagus was perfused with PBS. Tantalum disks were then injected into the trachea one at a time, and their position was captured and recorded at intervals during their movement orad from the digitized radiographic images of the trachea. The vagus nerves were then transected or atropine was administered, and tracheobronchial transport was tested again. The esophageal perfusate was changed to HCl for 30 min, and tantalum disks were again injected into the trachea one at a time and their rate of movement was recorded over the next 60 min.

BRONCHOTRACHEAL MUCOUS OUTPUT.

We perfused the airway with 0.9% NaCl at 0.25 ml/min for the entire experiment and collected mucous output every 30 min. For the first 60 min the esophagus was perfused with PBS: the first 30 min allowed for the system to stabilize, and the second 30 min was the control period for statistical comparison. The vagus nerves were then transected or atropine was administered, and the esophagus was perfused with PBS for another 30 min. The esophageal perfusate was changed to HCl for 30 min and then to PBS for 60 min.

Control of mucous production.

SELECTIVE PERFUSION.

The proximal or distal esophageal segment was perfused with HCl for 30 min and then with PBS for another 30 min. At the end of the experiment the trachea adjacent to the separate esophageal segments was removed and placed in 10% formalin for later histological analysis. The two separate tracheal segments were 1–3 cm and 5–7 cm from the cricoid cartilage.

Techniques

X-ray image system.

The microfocus X-ray system is composed of a FeinFocus-100 X-ray source (3 mm focal spot), an image intensifier (model AI-5830-HP, North American Imaging) coupled to a charge-coupled device (CCD) camera (Silicon Mountain Design), and a specimen micromanipulator stage (New England Affiliated Technologies), all mounted on a precision rail with position information provided by linear encoders (Mitutoyo). The geometry of the system allows magnification of the specimen to be adjusted by changing the specimen's proximity to the X-ray source. The image data were sent from the CCD camera to a frame grabber board mounted in a Dell 610 workstation running the Windows NT operating system.

Bronchial diameter measurement.

The lungs of the cat were visualized using the microfocus X-ray system as described previously (31). An appropriate lung field that contained airways of many different diameters and in which the tantalum had dusted the airways well enough to allow measurement of the diameters and evenly enough so that clumping of tantalum did not occur was selected. The X-ray stage and image intensifier were then fixed at this position, and radiographic images of the lung field were obtained every 2 min throughout the study. The diameters of various airways were quantified later (see below). The control values were obtained during the last 5 min of the initial 0.1 M PBS perfusion, because the preparation is well stabilized by this time.

Quantification of bronchial diameters.

The diameter of the airways was determined in an objective fashion as described previously (31) using mathematical models originally developed for quantification of pulmonary vasculature (9) and later adapted for the airways (55). One of two methods was used depending on the size and location of the airway. For small- and intermediate-diameter (<2.1 mm) airways, we used the mathematical model. This method is optimal when there is sufficient distance between airway segments of interest and there is a region of background on both sides of the airway. This is commonly the case for the intermediate and smaller airways. For large (>2.1 mm) airways, the diameter can be measured by manual placement of a line region of interest perpendicular to the axis from edge to edge of the airway.

The radiographic image is a two-dimensional projection of a three-dimensional object, i.e., the lungs, and objects closer to the X-ray source appear larger in the image. Therefore, each measurement of diameter in the radiographic image was calibrated separately. Distance within the plane of the radiographic image was calibrated by moving the cat a known distance perpendicular to the X-ray beam while the image was recorded. The distance moved in pixels of a shadow of an airway of interest in the image per micrometer of movement was the distance calibration factor for that airway.

Quantification of tracheal mucociliary transport rate.

Tracheal mucociliary transport rate was determined by acquiring images of the trachea while the injected tantalum disk traversed the field of view as described previously (31). This technique was adapted from similar techniques used by others (8, 62). We calculated the transport rate by allowing the disk to traverse as much of the field of view as possible and determining the time and distance traveled. The average velocity was calculated as the slope of the regression line of the distance vs. time of travel of each disk. The velocity of at least three disks for each experimental condition was determined and averaged to obtain the representative average value for that experimental condition.

Quantification of mucous output.

We quantified mucous secretion on the basis of concentration of a major constituent of mucus, rather than volume of fluid collected from the trachea, because in preliminary experiments we found that the volume of fluid collected from the trachea changed little during the experiments. Hexosamines are major constituents of mucus (26, 29), and hexosamine concentration of the perfusates was determined using a modification of a standard technique (59) we developed previously (32). Hexosamine output varied considerably among animals and across time; therefore, hexosamine output was quantified every 30 min and expressed as a percentage of the initial hexosamine output for each animal.

Quantification of tracheal mucous content.

Prior studies have suggested that quantitative histochemistry of the trachea should be done using longitudinal sections, rather than cross sections, because of the significant variation in the number of submucosal glands in the longitudinal direction (26). The majority of the tracheal submucosal glands are found between the cartilage rings (7). However, there is variability in the glands in cross, as well as longitudinal, sections. Most of the glands are located in the dorsolateral portion of the cross section (7), i.e., at 4 and 8 o′clock positions, with 12 o′clock being the most ventral point. Because our goal was to compare tracheal regions along the longitudinal axis, we took into consideration the variability in location of the submucosal glands in all planes. We sliced the trachea in cross section primarily between the cartilage rings, and we selected for quantification those cross sections with the most submucosal glands at their 4 and 8 o′clock positions. We quantified mucous content as a percentage of the total area of the epithelial layer, i.e., mucosa and submucosa. Therefore, our measure of mucous content optimized the sections with the most submucosal glands in all planes but also quantified mucus as a percentage of the total area of the epithelial layer to minimize variability among animals.

The fixed tracheal segments were embedded in paraffin and sliced at 5 μm and stained with Alcian blue (AB) at pH 1.0 and periodic acid-Schiff (PAS). The sections were then magnified ×10 (Olympus BH-2), and micrographs (Spot RT digital microscope camera) of the entire sections were taken (Fig. 1). Coverage of the entire tracheal section required 14–18 micrographs, and micrographs of 2 complete tracheal sections, selected as described above, were obtained for each experimental group. In the longitudinal direction of the trachea, the number of mucous cells is greatest between the cartilage rings; therefore, we targeted this region. Because the number of mucous cells varied longitudinally, we quantified all the mucous cells of two tracheal sections to control for sample bias. The results of the two sections were averaged.

Fig. 1.

Technique for separation of periodic acid-Schiff (PAS)- and Alcian blue (AB)-stained sections. The original image is a gray-scale image of the original PAS/AB-stained tracheal section. PAS and AB images are appropriately filtered (as described in materials and methods) original images that illustrate the type of images used to quantify relative amounts of PAS- and AB-stained mucus in tracheal sections.

Two types of mucous glands, mucosal and submucosal, are differentiated by anatomic position (Fig. 1), and two types of mucus, acidic and neutral, are differentiated by staining properties (Fig. 1): acidic mucus is stained by AB (blue color) and neutral mucus is stained by PAS (pink color). The areas of total, neutral, and acidic mucus were quantified for mucosal and submucosal mucous glands using ImageJ software.

TOTAL MUCUS.

ImageJ software was used to auto color-balance and then split the red-green-blue (RGB) color space of the image, such that green was used to quantify mucus within the submucosa, red was used to quantify mucus in the mucosa, and blue was discarded. The areas of mucus in each epithelial layer, i.e., mucosa and submucosa, were quantified using the area function, such that the area threshold just filled the mucous cells to capacity. These threshold values differed for each image and were saved for quantification of mucous type (see below). The percentage of mucus in each layer was calculated by dividing the area of mucus in each layer by the total area of that layer, i.e., mucosa or submucosa.

TYPE OF MUCUS.

The areas of tissue containing neutral (PAS-stained) and acidic (AB-stained) mucosubstances were separated by adjustment of the threshold of the hue of the RGB color space of the color-balanced image using the color threshold function. The threshold that maximally separated these areas was selected, and it averaged 187 ± 3 (n = 26) pixel intensity value (of a maximum of 256). The thresholded image was stop-filtered to create the AB-stained image (Fig. 1) and pass-filtered to create the PAS-stained image (Fig. 1). The areas of the mucosa and submucosa containing mucus of the thesholded images, i.e., AB- and PAS-stained, were then quantified as described above for total mucus using the same area thresholds. The percentage of each type of mucus, neutral (PAS-stained) and acidic (AB-stained), in each epithelial layer, i.e., mucosal or submucosal, was calculated.

Statistics

For studies of multiple experimental means, ANOVA was followed by a multiple-comparison test. For comparison over time, repeated-measures ANOVA was used; for comparison across time, one-way ANOVA was used. When the objective was to compare all groups with each other, Tukey's multiple-comparison test was used. For the histological studies, differences between the control and experimental areas were tested using Wilcoxon's matched-pair signed-rank test. Values are means ± SE. P ≤ 0.05 was considered statistically significant.

RESULTS

Effect of Vagotomy

Airway diameter.

After the vagus nerves (n = 7) were cut, the airways of all diameters significantly dilated (Figs. 2 and 3A), on average 35% (Fig. 3A), while the esophagus was perfused with PBS. No significant effects on airways of any diameter were observed for up to 60 min after the perfusate was changed to HCl (Figs. 2 and 3A).

Fig. 2.

Radiographic evidence of effects of vagotomy on airway diameter responses to esophageal acid exposure, shown as a series of radiographs of the airways of a cat with tantalum-dusted airways. Comparison of regions designated by arrows demonstrates that vagotomy increased airway diameter but blocked the airway diameter response to esophageal acidification.

Fig. 3.

Effects of vagotomy or atropine on airway diameter responses to esophageal acid exposure. Airways were grouped into 3 categories based on diameter and compared over time during exposure to different esophageal perfusates, i.e., PBS or HCl. A and B: effects of vagotomy and atropine on airway diameter response to esophageal HCl exposure. Either vagotomy or atropine significantly increased the diameter of airways of all 3 sizes, but esophageal acid exposure under either experimental treatment had no significant effect on airway diameter. *P < 0.05 vs. corresponding control (by repeated-measures ANOVA with Tukey's multiple-comparison test); n = 7 for each study.

Mucociliary transport.

After the vagus nerves (n = 5) were cut, there was no significant change in tracheal mucociliary transport rate (Fig. 4A) while the esophagus was perfused with PBS. However, after the perfusate was changed to HCl, the tracheal mucociliary transport rate was significantly reduced by ∼75% during the first 30 min and by 50% for up to 1 h after the start of HCl perfusion (Fig. 4A).

Fig. 4.

Effects of vagotomy or atropine on tracheal mucociliary transport rate response to esophageal acid exposure. A and B: effects of vagotomy and atropine on mucociliary transport response to esophageal HCl exposure. Vagotomy or atropine had no significant effect on tracheal mucociliary transport rate. On the other hand, esophageal acid exposure significantly decreased mucociliary transport after vagotomy, but not after atropine administration. *P < 0.05 vs. PBS before vagotomy; #P < 0.05 vs. PBS after vagotomy (by repeated-measures ANOVA with Tukey's multiple-comparison test); n = 5 for each study.

Mucous output.

We found that, compared with the control period of the first 30 min of PBS perfusion of the esophagus (n = 7), the basal level of hexosamine output continually declined over time (Fig. 5). When we statistically compared the effects over (Fig. 5) or across (Fig. 6) time, we found that hexosamine output significantly increased after esophageal acidification (Figs. 5 and 6; n = 7). Vagotomy (n = 8) did not significantly alter the hexosamine output response during the first 30 min of HCl administration (Fig. 6) but did significantly reduce the response to esophageal acidification from 30 to 90 min after HCl administration (Fig. 6).

Fig. 5.

Effects of vagotomy and atropine on tracheal mucous output stimulated by esophageal acid exposure: comparison within treatment. Change in mucous output is shown over 4 time periods during 4 different experimental treatments. All values represent percentage of the control measure obtained during the initial 0–30 min of perfusion of the esophagus with PBS. Experimental alteration began at the end of the 30-min control period. HCl was administered at 60–90 min. HCl significantly increased mucous output for the 1st h, and vagotomy blocked this response at all time periods, while atropine block at all time periods but the first was caused by esophageal acid exposure. *P < 0.05 for difference among time periods during each experimental treatment (by repeated-measures ANOVA with Tukey's multiple-comparison test).

Fig. 6.

Effects of vagotomy and atropine on airway mucous output stimulated by esophageal acid exposure: comparison across treatment. Change in mucous output across experimental treatments is shown during the same time period. Values are shown as percentage of the control measure obtained during the initial 0–30 min of perfusion of the esophagus with PBS. Experimental alteration was begun at the end of the 30-min control period. HCl was administered at 60 and 90 min. HCl significantly increased mucus output, and atropine blocked this response at all time periods. On the other hand, vagotomy did not block the initial mucus output caused by esophageal acidification but did block the delayed increase in mucous output. *P < 0.05 vs. PBS; #P < 0.05 vs. HCl (by ANOVA with Tukey's multiple-comparison test); n = 5, 7, 8, and 10 for PBS, HCl, HCl-vagx (HCl + vagotomy), and HCl-atr (HCl + atropine), respectively.

Effect of Atropine

Airway diameter.

After atropine administration (n = 7), airways of all diameters significantly dilated (Fig. 3B), on average, 32% while the esophagus was perfused with PBS. No significant effects on airways of any diameter were observed for up to 60 min after the perfusate was changed to HCl (Fig. 3B).

Mucociliary transport.

After atropine administration (n = 5), there was no significant change in tracheal mucociliary transport rate (Fig. 4B) while the esophagus was perfused with PBS. In addition, there was no significant change in tracheal mucociliary transport rate for up to 1 h after the perfusate was changed to HCl (Fig. 4B).

Mucous output.

Atropine (n = 10) caused no significant change in basal mucous output (Figs. 5), but it significantly reduced the increase in mucous output caused by esophageal acid exposure across all time periods (Fig. 6).

Control of Mucous Production

Selective perfusion.

We found that selective perfusion of a segment of the esophagus (n = 11), i.e., proximal or distal cervical esophagus, with HCl (pH 1.2) or PBS (pH 7.4) resulted in significantly more total mucous production in the submucosal glands of the trachea adjacent to the HCl- than PBS-perfused segment (Fig. 7). This relationship of tracheal mucous production to adjacent esophageal pH was also observed in the mucosa when HCl was perfused into the distal esophageal segment. When HCl was perfused into the proximal segment, no relationship with mucous production of the mucosa in the adjacent tracheal segment was observed (Fig. 7).

Fig. 7.

Effect of selective perfusion of proximal (Prox) and distal (Dist) esophageal segments on mucous production of the adjacent trachea, shown as percent area of the submucosa and mucosa of tracheal transections containing material stained with AB, PAS, or both (total). Tracheal transections were adjacent to the esophageal segments that contained PBS (pH 7.4) or HCl (pH 1.2) and were labeled according to the fluid to which the adjacent esophageal segment had been exposed. Whether the proximal or distal esophageal segment was perfused with HCl, mucous production was significantly greater in the trachea adjacent to the esophageal segment exposed to HCl than in the trachea adjacent to the esophageal segment exposed to PBS. The only exception was mucous production of the mucosa when the proximal esophageal segment was perfused with HCl. Therefore, mucous production of the trachea, especially of the submucosal glands, is highly related to the acid level of the corresponding adjacent esophagus. *P < 0.05 for adjacent regions of the trachea, each containing fluid of different pH (by Wilcoxon's matched-pair signed-rank test); n = 11.

Type of mucus.

We found that the primary type of mucus produced by the submucosal glands of the trachea is stained by PAS and that the primary type of mucus produced by the mucosal glands is stained with AB (Fig. 7). Perfusion of either segment, proximal or distal cervical, of the esophagus with HCl or PBS resulted in significantly more PAS-stained mucous production in the submucosal glands of the trachea adjacent to the HCl- than PBS-perfused esophageal segment (Fig. 7). The same relationship of tracheal mucous production to adjacent esophageal pH was also observed for the PAS-stained mucous production of the mucosal glands when HCl was perfused through the distal esophageal segment. The production of AB-stained mucus was not related to the pH of either segment for either the mucosal or submucosal glands except for one situation: when HCl was perfused through the distal cervical esophageal segment and PBS was perfused through the proximal esophageal segment, the PAS-stained mucous production of the mucosa was significantly greater in the trachea adjacent to the HCl- than the PBS-perfused distal segment (Fig. 7).

DISCUSSION

We found no airway diameter response to esophageal acidification after vagotomy or atropine, but we previously found (31) that esophageal acidification significantly decreased the diameter of the airways. Therefore, the airway constriction response to esophageal acidification is mediated by vagal cholinergic fibers.

This is consistent with our previous findings that the primary reflex pathway controlling airway smooth muscle (5, 6, 60), as well as mediating airway resistance (60) and airway diameter (60), is composed of vagal cholinergic fibers. In addition, vagal stimulation has been found to primarily constrict the smaller airways (4), which are reflexly constricted by esophageal acid exposure (31).

We also found no mucociliary transport response to esophageal acidification after atropine administration, but we previously found (31) that esophageal acidification significantly decreased mucociliary transport. Therefore, cholinergic receptors mediate the mucociliary transport response to esophageal acid exposure. In addition, we found that atropine blocked the mucous output response to esophageal acid exposure, indicating that the mucous response is also mediated by cholinergic receptors. These findings are consistent and support each other, as mucociliary transport rate is inversely related to mucous output (49, 54). This finding is supported by prior findings that tracheal mucous secretion activated reflexly (21, 25, 44), neurally (2, 3, 53), or by vagal stimulation (16, 16, 42, 47, 57) is mediated by cholinergic receptors.

We found that vagotomy did not block the initial effects of esophageal acidification on increasing mucous output or prevent a significant decrease in tracheal mucociliary transport rate caused by esophageal acid exposure. These findings are consistent with each other, as there is an inverse relationship between mucous output and mucociliary transport rate (49, 54). On the other hand, vagotomy significantly reduced mucous output after 30 min of esophageal acidification, when the mucociliary transport rate was still significantly reduced. This suggests that mucociliary transport rate and mucous output are not always inversely related and that the vagus nerves may affect this relationship. There are many possible explanations for these results, as there are many factors that control mucociliary transport rate and mucous output. Mucociliary transport rate is dependent on the viscosity and elasticity of the mucus lining the airway (30, 51), the depth of the mucous layer (51), the depth of the periciliary fluid layer (51), and ciliary beat frequency and velocity (10, 49, 51, 63). Mucous output depends on 1) the source of the mucus, goblet vs. apocrine gland cells, as the different sources have different constituents (17, 18, 29, 52), and 2) how the mucus is measured (2, 3, 11, 16–18, 29–31, 45, 47, 48, 52, 57), as there are many different constituents of mucus, which can be neurochemically altered (16–18, 29). Considering that we have not recorded all these factors, we cannot make conclusions about this observation; therefore, more studies are needed to resolve this issue.

Our finding that the vagal innervation had no role in the initial effects of esophageal acidification on mucous output seems to contradict previous reports that tracheal mucous output is mediated by the vagus nerves (2, 3, 16, 18, 21, 25, 42, 47, 50). However, our results do not contradict other studies; they show that there are more ways to activate mucous secretion than through the vagus nerves, as described below.

There are two sources of mucus in the airways, goblet cells of the mucosa and apocrine gland cells of the submucosa, but the apocrine gland cells outnumber the goblet cells (48). We found that the area of mucus in the mucosa was similar to or larger than that in the submucosa, but this was because the total area of the submucosa was larger than that of the mucosa. In cats the source of most secretions collected from the trachea in response to neural or chemical stimulation is primarily the submucosal, rather than goblet, cells (17, 18, 29, 52). In addition, in cat trachea (52) cholinergic agonists do not stimulate secretion from goblet cells. Therefore, it is likely that the source of most of the mucus collected in our studies was submucosal apocrine gland, rather than mucosal goblet, cells.

We found that the amount of mucus produced by the trachea was related to the pH of the adjacent esophagus. Therefore, it is likely that there is a local connection between the esophagus and adjacent trachea controlling this effect. We hypothesize that this local connection is mediated by the previously defined direct neural pathway from the esophagus to the trachea ganglia (35). Further studies are needed to confirm this relationship and to identify the local mediator.

We also found that the majority of the mucosubstances within the submucosal apocrine glands and activated by esophageal acid exposure are neutral, as they are mostly stained by PAS (20, 26, 36). On the other hand, the majority of the mucosubstances of the mucosal goblet cells are acid mucosubstances, as they are stained primarily by AB (20, 26, 36). Therefore, esophageal acid exposure is associated with production of neutral mucosubstances of the submucosal apocrine glands of the adjacent trachea. Acid mucosubstances are produced and secreted by mucosal cells in a position to provide defense against bacteria, e.g., the colon (45), whereas neutral mucosubstances are produced and secreted primarily by submucosal glands in a position to defend against acid, e.g., the duodenum (38, 46). Therefore, the increased production of PAS-stained mucus from submucosal glands of the trachea with acid exposure of the adjacent esophagus is the appropriate response to the possible threat of acid exposure of the airway.

The effect of acid perfusion on mucous production of the adjacent trachea seemed to be greater and more widespread for the distal segment of the esophagus, i.e., both mucosal and submucosal glands were affected, than for the proximal segment. This greater effect may have been due to the larger stimulus, as the distal segment encompassed a three-times-longer length of esophagus than the proximal segment.

A significant difference between our studies and those of others that investigated the role of extrinsic reflexes or vagal stimulation in the control of tracheal mucous output was the magnitude and duration of the responses. In our study the magnitude of the mucous response was relatively low, ≤150% of control, but the duration was long, ≥60 min after the stimulus. However, others stimulating extrinsic reflexes (25, 50) or the vagus nerves (16, 18, 42, 47) recorded mucous responses many times greater than the control response, which lasted not much longer than the duration of the stimulus. This dichotomy suggests that there may be two types of tracheal mucous secretions that are controlled differently.

Two basic types of responses of the airway submucosal glands have been reported (61): an innate mucosal defense system and an emergency airway defense system. The innate system primarily regulates housekeeping functions by secretion of mucus onto the airway surface to aid clearance and protection from physiological levels and types of offending substances, such as bacteria. Studies (1, 18) have found that the resting tracheal mucous output is not vagally dependent, and because this is the resting mucous output, it is likely to be part of the innate defense system. The innate system is largely locally controlled, and its responses are moderate. On the other hand, the emergency system reacts to remove and neutralize nonphysiological destructive substances, such as foreign bodies. The emergency response is large and extrinsically controlled. Therefore, based on our findings in this study, we conclude that the tracheal mucous response to acid exposure of the esophagus is primarily part of the airway innate defense system controlling housekeeping functions, and not part of an extrinsic emergency response. This conclusion is also consistent with our finding that esophageal acid exposure caused reflex contraction of the smaller, rather than larger, airways, causing no change in airway resistance.

The nature of the airway response to esophageal acid exposure as part of the innate defense system is not only consistent with our results, but it is consistent with the nature of esophageal acid exposure. In normal individuals, the esophagus is often exposed to gastric reflux (41), and microaspiration of gastric juice occurs in normal individuals (40). In addition, microaspiration has been associated with a number of airway symptoms (27) and diseases (21, 56). Therefore, just as the airway must provide a constant defense system to bacteria, it must also provide a constant defense system to acid. While mucosubstances produced and secreted by the mucosal cells are particularly effective against bacteria (33, 45), mucosubstances produced and secreted by submucosal glands are particularly effective against luminal acid (12, 22).

Limitiations

Control experiments on the effects of esophageal acid exposure on airway diameter or mucociliary transport, which were conducted previously (31), were not conducted for this study, because repeating them would violate the principle of the Animal Welfare Act regarding minimization of the number of animals used for experimentation. However, we did conduct control studies in experiments that were significantly different from our prior study, i.e., mucous output, or in experiments that were new, i.e., mucous production. The mucous output study was different, in that the cats were placed at a 10° angle head-down to assist in the collection of mucus. However, this resulted in a stronger stimulation, because gravity caused the acid to remain longer in the cervical region of the esophagus, and the result was a longer-lasting mucous response.

In summary, esophageal acidification causes four airway responses: 1) airway constriction primarily of the smaller-diameter airways, 2) increased mucous, i.e., hexosamine, output, 3) deceased bronchotracheal mucociliary transport rate, and 4) increased mucous production, primarily by the apocrine glands. Our data suggest that acid exposure of the esophagus causes reflex constriction of the airway through a vagal cholinergic pathway and reflex inhibition of tracheal mucociliary transport through a nonvagal cholinergic pathway. The tracheal mucous output response to esophageal acid exposure is partly mediated by an immediate nonvagal cholinergic pathway and a delayed vagal cholinergic pathway. Also, stimulation of production of neutral mucosubstances by the apocrine gland cells, possibly through a local neural pathway, is delayed. Considering the delayed nature and the low magnitude of the mucous responses, the acid-buffering capacity of apocrine gland secretion, and the local neural control of the apocrine gland response, we hypothesize that the airway responses to esophageal acidification are part of the innate, rather than acute emergency, defense system of the airway.

GRANTS

This work was supported in part by National Institutes of Health Grants P01-DK-068051, RO1-DK-25731, and HL-19298, the W. M. Keck Foundation, and the Department of Veterans Affairs.

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

I.M.L. and R.S. developed the concept and designed the research; I.M.L., S.T.H., and B.K.M. performed the experiments; I.M.L. and S.T.H. the analyzed data; I.M.L. and S.T.H. interpreted the results of the experiments; I.M.L. and S.T.H. prepared the figures; I.M.L. drafted the manuscript; I.M.L., H.V.F., and R.S. edited and revised the manuscript; I.M.L., S.T.H., B.K.M., H.V.F., and R.S. approved the final version of the manuscript.