A morphometric analysis of the autonomic innervation of cat tracheal glands

Airway Gland Structure and Function

Submucosal glands contribute to airway surface liquid (ASL), a film that protects all airway surfaces. Glandular mucus comprises electrolytes, water, the gel-forming mucin MUC5B, and hundreds of different proteins with diverse protective functions. Gland volume per unit area of mucosal surface correlates positively with impaction rate of inhaled particles. In human main bronchi, the volume of the glands is ∼ 50 times that of surface goblet cells, but the glands diminish in size and frequency distally. ASL and its trapped particles are removed from the airways by mucociliary transport. Airway glands have a tubuloacinar structure, with a single terminal duct, a nonciliated collecting duct, then branching secretory tubules lined with mucous cells and ending in serous acini. They allow for a massive increase in numbers of mucus-producing cells without replacing surface ciliated cells. Active secretion of Cl(-) and HCO3 (-) by serous cells produces most of the fluid of gland secretions. Glands are densely innervated by tonically active, mutually excitatory airway intrinsic neurons. Most gland mucus is secreted constitutively in vivo, with large, transient increases produced by emergency reflex drive from the vagus. Elevations of [cAMP]i and [Ca(2+)]i coordinate electrolyte and macromolecular secretion and probably occur together for baseline activity in vivo, with cholinergic elevation of [Ca(2+)]i being mainly responsive for transient increases in secretion. Altered submucosal gland function contributes to the pathology of all obstructive diseases, but is an early stage of pathogenesis only in cystic fibrosis.

Mucus secretion from individual submucosal glands of the ferret trachea

Mucus secretion from individual tracheal glands in adult ferrets was studied with time-lapse optical imaging of mucus droplets under an oil layer. Density of functional glands (determined by responses to 1 muM carbachol) was 1.5 +/- 0.3 per mm(2) (n = 6). Secretion rates (in pl.min(-1).gland(-1)) were as follows: 4.1 +/- 0.7 basal (unstimulated; n = 27, 669 glands), 338 +/- 70 to 10 microM forskolin (n = 8, 90 glands), 234 +/- 13 to 1 microM VIP (n = 6, 57 glands), 183 +/- 92 to 10 microM isoproterenol (n = 3, 33 glands), 978 +/- 145 to 1 microM carbachol (n = 11, 131 glands), and 1,348 +/- 325 to 10 muM phenylephrine (n = 7, 74 glands). The potency (EC(50), in microM) and efficacy (V(max), in pl x min(-1) x gland(-1)) were 7.6 (EC(50)) and 338 +/- 16 (V(max)) to forskolin, 1.0 (EC(50)) and 479 +/- 19 (V(max)) to VIP, 0.6 (EC(50)) and 1,817 +/- 268 (V(max)) to carbachol, and 3.7 (EC(50)) and 1,801 +/- 95 (V(max)) to phenylephrine. Although carbachol and phenylephrine were equally effective secretagogues, only carbachol caused contractions of the trachealis muscle. Synergy was demonstrated between 300 nM isoproterenol and 100 nM carbachol, which, when combined, produced a secretion rate almost fourfold greater than predicted from their additive effect. The dependence of fluid secretion on Cl(-) and HCO(3)(-) varied depending on the mode of stimulation. Secretion stimulated by VIP or forskolin was reduced by approximately 60% by blocking either anion, while carbachol-stimulated secretion was blocked 68% by bumetanide and only 32% by HEPES replacement of HCO(3)(-). These results provide parametric data for comparison with fluid secretion from glands in ferrets lacking CFTR.

Regulation of secretion from serous and mucous cells in the trachea

The physical properties of mucus and the efficiency of tracheal mucociliary clearance depend on maintenance of a balanced interaction among several epithelial cell types. Some of these cell types are specialized to perform ion and water transport, others to perform synthesis and secretion of macromolecules. Our studies have been aimed specifically at identifying the neural mechanisms regulating macromolecule secretion from two of these cell types, i.e. serous and mucous gland cells. Because these cells occur as part of a complex epithelium, it is difficult to monitor the properties and functions of each cell type individually. We have therefore relied principally on morphological methods, which can potentially focus on a single cell type within a heterogeneous tissue. Such studies, however, depend on the availability of visible markers (enzyme-labelled antibodies, radioligands, etc.), and many important aspects of gland cell function cannot be assessed morphologically. Two alternative approaches are therefore being developed: the isolation and segregation of gland cells according to type, and the production of monoclonal antibodies that recognize secretory products of individual cell types. These methods allow serous and mucous cells to be studied by biochemical as well as morphological methods.

Precise localization of VIP-, NPY-, and TH-immunoreactivities of cat laryngeal glands

The precise distribution of vasoactive intestinal polypeptide (VIP)-, neuropeptide Y (NPY)-, and tyrosine hydroxylase (TH)-immunoreactive (ir) fibers in the cat's laryngeal glands was examined by immunoelectronmicroscopy. A relatively dense population of VIP-ir fibers was recognized close to the basal lamina of the glandular and myoepithelial cells. Some VIP-ir fibers contacted with the basal lamina and some of them pierced it and ran intercellularly in the adjoining glandular cells without making synaptic contacts with them. NPY- and TH-ir fibers were located in the vicinity of the basal lamina, but they were less abundant than VIP-ir fibers at this region. They never terminated or penetrated the basal lamina. Pattern of distribution of TH-ir fibers was similar to that of NPY-ir fibers. The estimated ratio of VIP-, TH-, and NPY-ir fibers was 20:4:1 from the density of fibers in the laryngeal glands. This value was equal between serous and mucous glandular cells, so both types of glandular cells may receive the same pattern of autonomic innervation.

Distribution of substance P immunoreactive nerve fibers in the tracheal submucosal gland of cats

Immunohistochemistry combined with electron microscopy was employed to investigate the distribution of substance P-immunoreactive (SP-IR) nerve fibers in the tracheal submucosal gland of cats. The SP-IR nerve fibers were found to form a network around the glands. Numerous varicosities were also detected within the basement membrane of the acini and secretory tubules. All the intraglandular varicosities showed close spatial contact with serous cells, mucous cells, and myoepithelial cells. Our findings suggest that substance P-induced mucus secretion from tracheal submucosal glands in cats may be caused not only by a glandular contractile response of myoepithelial cells, but also by direct stimulation to both serous and mucous cells.

Ultrastructural localization and identification of adrenergic and cholinergic nerve terminals in the olfactory mucosa

Pharmacological and ultrastructural methods were used to demonstrate alpha-adrenergic regulation of secretory granule content of acinar cells of Bowman's glands and to localize and identify adrenergic and cholinergic axonal varicosities and terminals in the olfactory mucosa of the tiger salamander. The alpha-adrenergic agonist phenylephrine caused secretory granule depletion from Bowman's glands; the alpha-adrenergic antagonist phentolamine partially blocked this effect. These observations were quantified using light microscopic computer-assisted morphometric techniques. Both drugs caused morphological signs of electrolye/water transport. Adrenergic axonal varicosities were identified by the presence of small granular vesicles (SGVs, 45-60 nm in diameter) containing electron-dense material that was enhanced by 5-hydroxydopamine loading and chromaffin reaction fixation techniques. Throughout the lamina propria, small fascicles with axons containing SGVs as well as varicosities and terminals with SGVs were located adjacent to blood vessels, Bowman's gland acini, and melanocytes. Mean vesicle diameters at these sites were 54 +/- 7 nm, 50 +/- 9 nm, and 56 +/- 8 nm, respectively; varicosities were located approximately 0.1-1.0 microns from their presumed cellular targets. Axonal varicosities containing small agranular vesicles (AGVs, 65 +/- 8 nm in diameter), identified as cholinergic by their size and by the absence of electron-dense material after 5-hydroxydopamine loading and chromaffin reaction fixation, were located between adjacent acinar cells. In addition, adrenergic varicosities containing SGVs (56 +/- 6 nm in diameter) were found within 1 micron of blood vessels associated with Bowman's gland ducts and sustentacular cells near the base of the olfactory epithelium. These results characterize the ultrastructural basis for adrenergic and cholinergic regulation of vasomotor tone and secretion within the olfactory mucosa.

Characterization of beta-adrenergic receptors in cultured bovine tracheal gland cells

We characterized the beta-adrenergic receptors that mediate secretory responses to isoproterenol in cultured bovine tracheal submucosal gland cells. Previous studies have shown that these cells have morphological and biochemical features characteristic of serous cells. Isoproterenol, epinephrine, and norepinephrine each stimulated the secretion of 35SO4-labeled macromolecules from these cultured serous cells with a rank order of potency (isoproterenol greater than epinephrine greater than norepinephrine) consistent with the presence of beta 2-adrenergic receptors. These functional studies were supported by radioligand-binding studies using [I125]-iodocyanopindolol (125I-CYP) to identify beta-adrenergic receptors. 125I-CYP binding to membrane particulates prepared from cultured serous cells was saturable and of high affinity (equilibrium dissociation constant 20 +/- 3 pM; mean +/- SE, n = 6) and was antagonized stereoselectively by propranolol. Adrenergic agonists competed for 125I-CYP-binding sites with a rank order of potency characteristic of the beta 2-adrenergic receptor subtype. A specific beta 2-adrenergic receptor antagonist, ICI 118.551, competed for a single class of 125I-CYP-binding sites with high affinity (inhibition constant 1.8 +/- 0.3 nM, n = 3). We concluded that the secretory response of cultured tracheal gland cells to isoproterenol is a response mediated by beta-adrenergic receptors of the beta 2 subtype.

Tracheal gland mucous cells stimulated in vitro with adrenergic and cholinergic drugs

To determine the responsiveness of tracheal mucous cells to adrenergic and cholinergic stimulation, we analyzed changes in their structure induced by neurotransmitter-like agonists. Ferret tracheal rings were exposed for 30 min in vitro to one of the following: phenylephrine, isoproterenol, or bethanechol (all at 10(-5) M), in the presence of absence of appropriate antagonists. Electron microscopy and morphometric analysis revealed that the volume density of mucous cells (Vvmc, i.e. the space occupied by mucous cells in the submucosa) significantly decreased, and the surface density of mucous cell apical membrane (Svam) increased in response to isoproterenol and bethanechol but not to phenylephrine. In metabolic labeling experiments, the morphological changes were accompanied by secretagogue-evoked release of 35S-labeled macromolecules. Taken together, these data suggest that tracheal mucous cells secrete 35S-labeled macromolecules in response to beta-adrenergic and muscarinic agonists by an exocytotic process that involves a reduction in cell size.

The functional significance of the sympathetic innervation of mucous glands in the bronchi of man

1. Pieces of human bronchi, from lung resected for carcinoma of the bronchus, were mounted in Ussing chambers and given [35S]sulphate as radiolabelled precursor of mucous glycoproteins (mucins). The release of 35S, bound to macromolecules, into the luminal half-chamber was used as an index of mucin secretion. 2. Noradrenaline, at concentrations of 1, 10 and 100 microM, was given into both halves of the Ussing chamber. At the lowest concentration, noradrenaline failed to change mucin output, but at the two higher concentrations it stimulated output. 3. In other experiments the sympathetic nerves in the bronchial wall were labelled with 5-hydroxydopamine and examined under the electron microscope. The distances between adrenergic nerve varicosities and submucosal glands were measured; some sympathetic nerve varicosities were seen within 1 microns of gland cells. 4. A simple mathematical model for the diffusion of noradrenaline was used to predict the concentrations of the transmitter likely to result at different distances from a nerve if one or more vesicles of noradrenaline were released. 5. The model predicts that the release of a single large vesicle of noradrenaline is likely to generate an effective concentration of transmitter provided that the nerve is within 1 micron of the target cell.

Nervous control of mucin secretion into human bronchi

Pieces of ferret trachea and human bronchi were mounted in Ussing chambers and given [35S]sulphate as a radiolabelled precursor of mucous glycoproteins (mucins). The output of 35S bound to macromolecules was studied as an index of mucin secretion. In the ferret trachea, electrical field stimulation increased the rate of mucin secretion. Tetrodotoxin (10(-7) M or 10(-6) M) abolished this effect. Pilocarpine (25 microM) stimulated the output of mucins from human bronchus. Atropine (10(-5) M) abolished this effect. Electrical field stimulation of human bronchus stimulated mucin secretion. Tetrodotoxin (10(-6) M) abolished this effect. Field stimulation in the presence of either atropine (10(-5) M) or atropine with l-propranolol (10(-5) M) and phentolamine (10(-5) M) caused no stimulation of mucin secretion rate. Some bronchi were treated with noradrenaline (10(-5) M) for 1 h to allow the adrenergic nerves to take up transmitter. Even in these, atropine prevented the effect of field stimulation. We conclude that activity in cholinergic nerves can stimulate mucin secretion in the bronchi in man. Our results provide no evidence that the adrenergic nerves or non-adrenergic, non-cholinergic nerves have a direct action on bronchial secretory cells in man.

Control of mucoglycoprotein output from the rabbit nose

A method for the collection of rabbit nasal washings to analyse outputs of mucous glycoproteins is described. The radiolabels sodium [35S]sulphate and [3H]glucose are bound to the glycoproteins. The release of bound 35S and 3H was enhanced by cervical sympathetic nerve stimulation. Adrenoceptor agonists (phenylephrine, dobutamine and salbutamol) given I.V. increased the output of 35S, and the last two drugs increased the output of 3H. The blocking effects of thymoxamine and propranolol on these responses are described. Pilocarpine (given I.V. or intranasally) produced large increases in 35S release; histamine had little effect. Irritants (ammonia and cigarette smoke) and diluted serum or plasma, given intranasally, produced large increases in 3H output, and sometimes enhanced 35S release. Radiolabelled nasal washings fractionated on Sepharose CL-4B gel chromatography columns formed two peaks, with most of the radioactivity in the high molecular weight (mucous glycoprotein) peak.

Pig bronchial mucous membrane: a model system for assessing respiratory mucus release in vitro

A convenient organ culture system is described in which fragments of mucous membrane isolated from bronchi of the pig were maintained in either screw-cap tubes or multiwell tissue culture plates. The mucous membrane of the pig bronchus, like that of the human, is rich in mucus-secreting submucosal glands and can respond to cholinergic stimulation in vitro by releasing either L-[3H]fucose- or L-[3H]serine-labeled acid-precipitable macromolecules. Reproducible cholinergic-mediated release of labeled macromolecules was attained by first washing the mucous membrane fragments in serum-free modified Earles medium (Dulbecco's) for 120 min at 4 degrees C. Maximum stimulation was obtained when the incubation medium was supplemented with 0.5-2.0% horse serum. Approximately 50% of L-[3H]fucose-labeled macromolecules were eluted in the void volume from a column of Sepharose CL-6B in 6 M urea. Cochromatography of L-[3H]- and L-[14C]fucose-labeled glycoproteins released by mucous membranes of control and methacholine-treated tissue fragments failed to reveal any significant difference in any specific population of fucose-labeled glycoproteins. It is concluded that, as a whole, many different labeled molecules are released in response to cholinergic stimulation. Taken together, these results suggest that the mucous membrane of the porcine bronchus is a useful in vitro model for studying respiratory mucus secretion.

Mapping of adrenergic receptors in the trachea by autoradiography

We investigated the distribution of adrenergic receptors in ferret trachea using autoradiography. [3H]Dihydroalprenolol, used to identify beta-adrenoceptors, revealed a high density of specific binding sites over surface epithelium and submucosal glands, with less labelling of smooth muscle. [3H]prazosin labelling showed that alpha 1-receptors were numerous in glands and epithelium, but sparse in smooth muscle. Comparison of adrenergic receptor densities in tracheal sections from the same animals showed a rank order for submucosal glands of alpha 1 greater than beta, for epithelium beta greater than alpha 1 and for smooth muscle beta greater than alpha 1. Within the submucosal glands, alpha 1- and beta-adrenergic receptors were differentially distributed, with alpha 1-receptors being significantly more numerous over serous than mucous cells and beta receptors being significantly more numerous over mucous than serous cells. This technique provides insight into adrenergic regulation of airway function and should be useful in investigations of how relative receptor densities may be altered in disease.

Evidence for non-cholinergic, non-adrenergic nervous control of mucus secretion into the cat trachea

We have investigated the effects of electrical stimulation of the vagus nerves on the output of mucus glycoproteins (mucins), radiolabelled with 3H and 35S, into the trachea of anaesthetized cats. In five control experiments, stimulation of the vagus nerves on four successive occasions, separated by 1 h, caused significant rises in the output of radiolabelled mucins. In these experiments repetition of stimulation did not appear to lessen the response. In a parallel series of five experiments the vagus nerves were again stimulated on four occasions, but atropine was administered in increasing doses between the stimuli. Large responses, not significantly less than those seen in the corresponding control stimulations, were seen even in the presence of the highest dose of atropine. In this series of experiments, however, the effect of the last vagal stimulation (with the highest dose of atropine) was significantly less then the first (no atropine). Administration of phentolamine and l-propranolol in addition to atropine failed to reduce the response to vagal stimulation significantly. We conclude that, while cholinergic nerves can probably explain part of the increase in mucin output which occurs with vagus nerve stimulation, there is a large response mediated by a non-cholinergic, non-adrenergic neurotransmitter. Possible neurotransmitters and the relationship of these findings to those of earlier studies are discussed.

The control of mucin secretion into the lumen of the cat trachea by alpha- and beta-adrenoceptors, and their relative involvement during sympathetic nerve stimulation

We have tested the effects of phenylephrine, dobutamine and salbutamol, alpha-, beta 1- and beta 2-adrenoceptor agonists respectively, on the output of radiolabelled mucins into the cat trachea in situ. Phenylephrine significantly increased mucin output, an effect inhibited by the alpha-adrenoceptor antagonists, thymoxamine or prazosin, but not by propranolol. Dobutamine increased the output of 35S-labelled mucins greatly and had a smaller effect on 3H-labelled mucins. Propranolol blocked these effects but thymoxamine did not. At high doses atenolol, a beta 1-adrenoceptor antagonist, inhibited dobutamine's effect on 35S-labelled mucins. Salbutamol caused a small increase in mucin output and propranolol blocked this increase. Electrical stimulation of the sympathetic nerve supply to the trachea increased mucin output. Propranolol inhibited this effect; thymoxamine did not. We conclude that both alpha- and beta-adrenoceptors increase mucus secretion into the cat trachea but that only the beta-adrenoceptors respond to sympathetic nerve stimulation.

Ultrastructural and fluorescence histochemical studies on the sympathetic innervation of the canine laryngeal glands

The existence of the adrenergic terminals of the canine laryngeal glands has been revealed by electron microscopy and fluorescence histochemistry. Adrenergic fibres with fluorescent varicosities were observed around the base of th acini in the submucosa. In dogs treated with 5-OHDA, varicosities containing both small and large, dense-cored vesicles containing both small and large, dense-cored vesicles, believed to be adrenergic terminals, were found near blood vessels, gland cells and myoepithelial cells in the submucosal gland region. The role of these adrenergic terminals is briefly discussed.

Tracheal submucosal gland serous cells stimulated in vitro with adrenergic and cholinergic agonists. A morphometric study

A morphometric analysis was made of alterations in serous cell structure induced by adrenergic and cholinergic agonists. Ferret tracheal rings were exposed for 30 min in vitro to one of the following agonists: phenylephrine, terbutaline, or methacholine (all at 10(-5) M). Controls were incubated similarly in medium containing no drugs or medium containing both the agonist and an excess of the appropriate antagonist (phentolamine, propranolol or atropine, all at 10(-4) M). Electron microscopic observation and stereological analysis of the incubated samples revealed that the volume density of serous cell granules in controls (0.30 +/- 0.02, mean +/- SE, n = 4) was significantly reduced by phenylephrine (0.19 +/- 0.03, n = 4) and methacholine (0.17 +/- 0.01, n = 4), but not by terbutaline (0.27 +/- 0.04, n = 4). The presence of antagonists in the medium prevented the observed changes (phenylephrine/phentolamine: 0.29 +/- 0.03, n = 3 and methacholine/atropine: 0.33 +/- 0.06, n = 3). In addition, the volume density of intracellular vacuoles in controls (0.02 +/- 0.05, n = 4) was increased in response to methacholine stimulation (0.12 +/- 0.05, n = 4), but not in response to the other agonists. This effect was blocked by atropine (0.01 +/- 0.00, n = 3). We conclude that serous-cell granules are discharged by both alpha-adrenergic and cholinergic, but not beta-adrenergic stimulation. In addition, cholinergic stimulation evokes the formation of intracellular vacuoles, a possible indication of active ion and water transport.