INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: parasympathetic; autonomic; nonadrenergic, noncholinergic (NANC); reflex bronchospasm; reflex bronchodilitation

Airway smooth muscle tone is regulated primarily by the parasympathetic nervous system. Postganglionic parasympathetic nerves mediate both cholinergic contractions and nonadrenergic, noncholinergic (NANC) relaxations of airway smooth muscle. The opposing contractile and relaxant effects of these nerves ultimately determine airway caliber. Dysfunction of the parasympathetic nerves innervating the airways likely contributes to the pathogenesis of airways disease (1-3).

It had long been assumed that noncholinergic neurotransmitters were coreleased with acetylcholine from a single population of postganglionic parasympathetic nerves innervating the airways (3). Studies have shown, however, that parasympathetic nerve-mediated contractions and relaxations are probably mediated by distinct postganglionic nerves. In guinea pigs, for example, neurons intimately associated with the airway wall mediate cholinergic contractions. Conversely, neurons in the myenteric plexus of the adjacent esophagus, immunoreactive for nitric oxide synthase, and vasoactive intestinal peptide (and related peptides) provide relaxant innervation to airway smooth muscle (4-9). Electrophysiological studies indicate that the conduction velocities (and thus, likely, the myelination) of the preganglionic fibers innervating contractile and relaxant ganglia in guinea pigs may also be different (4). These observations suggest that anatomically and physiologically distinct vagal pathways mediate contractile and relaxant effects in the airways. Circumstantial as well as direct immunohistochemical and physiologic evidence suggests that the parasympathetic innervation of the airways of cats, ferrets, and humans may be comparable to that described in guinea pigs (3, 10-16).

Given the anatomic, physiologic, and neurochemical distinctions between these two parasympathetic pathways, differential regulation of contractile and relaxant reflexes in the airways seems plausible. Such potential for complexity of airway autonomic control has important implications relating to the role of autonomic nerves in health and disease. To address this issue, we compared contractile and relaxant reflexes in the airways of anesthetized guinea pigs in vivo, following selective activation of airway afferent nerve subtypes or extrapulmonary afferent nerves thought to regulate airway parasympathetic nerve activity. The results suggest that cholinergic and noncholinergic airway parasympathetic nerves are susceptible to differential reflex regulation.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal Preparation

Experiments described here were approved by the Johns Hopkins Medical Institutions Animal Care and Use Committee. Male Hartley guinea pigs (300-400 g; Hilltop, Scottdale, PA) were anesthetized with urethane (1 g/kg, intraperitoneally) and mechanically ventilated (60 breaths per minutes, 6 ml/kg, 2-3 cm H₂0 of positive end-expiratory pressure) following paralysis (2 mg/kg succinylcholine chloride subcutaneously). The abdominal aorta and vena cava were cannulated to monitor arterial blood pressure and for intravenous drug injection. Anesthesia was monitored throughout experiments by assessing cardiovascular changes to a sharp pinch of the animals paw.

Tracheal smooth muscle tension was measured in situ as described previously (17-20). The lumen of the tracheal segment studied was perfused with warmed (37° C), oxygenated Krebs buffer containing 3 µM indomethacin, 2 µM propranolol, and 1 µM phentolamine to block the effects of prostaglandins and circulating or neurally released catecholamines. Animals were also pretreated intravenously with propranolol (1 mg/kg) at the beginning of each experiment.

At the conclusion of each experiment, animals were killed by inhalation of 100% CO₂, delivered through the inspiratory port of the ventilator.

Alterations of Guinea Pig Tracheal Smooth Muscle Tone Evoked by Reflex or by Vagus Nerve Stimulation

Vagally mediated contractions and relaxations of the guinea pig trachealis were studied as described previously (17, 20). Animals were pretreated with methoctramine (1 mg/kg, intraveneously) to reduce the bradycardia associated with vagus nerve stimulation. Vagi were cut and placed on platinum electrodes attached to a Grass stimulator. Contractions were studied before any pretreatments. Relaxations were studied following pretreatment with 1 µM atropine and contraction of the trachealis with 10 µM histamine. Cumulative frequency- response curves were constructed (0.5-64 Hz) using optimal stimulation intensities (10-20 V, 1-millisecond pulse duration).

Reflex-mediated alterations in tracheal tone were then studied. Reflexes were initiated by a variety of stimuli: (1) intravenous injections of histamine (10 µg/kg [18]); (2) 1-30 µM capsaicin or 30 µM histamine applied to the laryngeal mucosa; (3) nebulized histamine (10 µg/ml), or capsaicin (100 µM); (4) increasing positive end-expiratory pressure (from 3 to 8 cm/H₂O); (5) a 1-2-minute period of ventilation with 100% N₂ (hypoxia, to stimulate chemoreceptors [21]); (6) application of cold (4° C) saline onto the nose (to evoke the diving reflex [22]); and (7) electrical stimulations of the phrenic or sub-bronchial vagus nerves (5-10 Hz, 10-20 volts) to evoke steady-state responses (typically 0.5-2 minutes).

When contractile and relaxant reflexes were initiated by histamine, the histamine H₁ receptor antagonist pyrilamine (1 µM) was added to the tracheal perfusate to block any direct effects of the autacoid on tracheal tone (18). This concentration of pyrilamine has no effect on muscarinic receptor function (23). For studying histamine-mediated relaxant responses, 1 µM PGF₂ (rather than 10 µM histamine) was used to contract the trachealis. Reflex-mediated alterations in cholinergic tone were expressed as a percentage change in cholinergic tone, a percentage of the maximum attainable contraction evoked by 300 mM BaCl₂, or as a percentage reversal of the histamine contraction where appropriate.

The reflex and parasympathetic nature of contractile responses evoked was assessed by either cutting both vagus nerves adjacent to the extrathoracic trachea (bilateral vagotomy), by the addition of the muscarinic receptor antagonist, atropine (1 µM) to the tracheal perfusate, or by ganglionic blockade with trimethaphan (5 mg/kg, intraveneously).

Data Analysis

All data are presented as mean ± SEM. Differences between group means were assessed using analysis of variance (ANOVA) followed by Scheffe's F test for unplanned comparisons (Statview; Berkely, CA). p values of < 0.05 were considered significant.

Drugs

Suppliers of drugs were as follows: Methoctramine hydrochloride (Tocris, Ballwin, MO); PGF₂ (Biomol, Plymouth, PA); Trimethaphan camsylate (Roche Laboratories, Nutley, NJ); all others (Sigma, St. Louis, MO). Drugs added to the tracheal perfusate were made in distilled water except: PGF₂ and indomethacin were dissolved in ethanol. Drugs administered intravenously, subcutaneously, or as an aerosol were dissolved in saline.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Contractions and Relaxations of the Trachealis Evoked by Electrical Stimulation of the Vagus Nerves In Situ

Bilateral electrical stimulation of the vagus nerves at optimal stimulation intensities (10-20 V, 1-millisecond pulse duration) evoked frequency-dependent, atropine-sensitive contractions of the guinea pig trachealis in situ (Figure 1). The mean ± SEM stimulation frequency that evoked contractions equivalent to 50% of the maximum contraction evoked by vagus nerve stimulation (EF₅₀) was 2.6 ± 0.9 Hz (n = 4). Following 1 µM of atropine administration and precontraction of the trachealis with 10 µM of histamine, vagus nerve stimulation also evoked frequency-dependent NANC relaxations of the precontracted trachea. The relaxant responses required higher stimulation frequencies than were needed to induce cholinergic contractions (EF₅₀, 5.8 ± 1.8 Hz, n = 4). Moreover, unlike the contractile responses, which began immediately upon nerve stimulation, reached equilibrium within 30-120 seconds at all stimulation frequencies studied, and rapidly reversed, NANC relaxant responses began 3-6 seconds after starting nerve stimulation, took 2-5 minutes to reach equilibrium, and slowly (3-6 minutes) recovered.

View larger version (20K): [in this window] [in a new window]   Figure 1.   Mean data for frequency-dependent, vagally mediated contractions (closed circles) and relaxations (open circles) of guinea pig trachea in situ. Data are presented as a percentage of the maximum response (contraction or relaxation) evoked by nerve stimulation. Vagi were cut and stimulated bilaterally on their peripheral ends using optimal stimulation intensities (10-20 V, 1-millisecond pulse duration). Frequency response curves (0.1-64 Hz) were generated cumulatively. Stimulation frequency was doubled when responses at a given frequency reached equilibrium or when 1-2 minute stimulation failed to evoke a response. Contractions were studied before any pretreatment. Relaxations were studied following administration of 1 µM of atropine to the tracheal perfusate and precontraction of the trachealis with 10 µM of histamine. Peak contractile effects evoked by vagus nerve stimulation averaged 57 ± 10% of the maximum contraction (n = 4). Peak relaxant effects evoked by vagus nerve stimulation reversed the histamine contraction by 45 ± 15% (n = 4). Atropine abolished vagally mediated contractions of the trachealis. The ganglionic blocker trimethaphan (5 mg/kg, intravenously) abolished both vagally mediated contractions and relaxations of the trachealis.

Vagus nerve stimulation also induced increases in pulmonary insufflation pressure (PT). Consistent with the widely held notion that this measurement is a particularly blunt tool for monitoring airway smooth muscle tone in vivo, the effects of vagus nerve stimulation on PT were all but unmeasurable (< 20% increase) at frequencies  2 Hz. The ganglionic blocker trimethaphan (5 mg/kg, intraveneously) markedly inhibited or abolished these vagally mediated effects on PT. Trimethaphan also abolished the contractile and relaxant effects of vagus nerve stimulation on tracheal tone, thus confirming the parasympathetic nature of these vagally mediated responses.

Reflex Alterations in Tracheal Tension Evoked by Intravenous Histamine

With the vagus nerves intact and before any pharmacological manipulations or challenges, baseline cholinergic tone averaged 33 ± 2% of the maximum contraction in the preparations used in this study (n = 48). Following intratracheal administration of the H₁ receptor antagonist pyrilamine (1 µM) to block the local effects of histamine on the tracheal segment studied, intravenous administration of 10 µg/kg histamine produced marked increases in cholinergic tone (53 ± 7% increase in cholinergic tone; Figure 2). Consistent with previous studies (18, 19), tracheal contractions evoked by histamine coincided with an increased PT (137 ± 32% increase above baseline) and an inconsistent effect on arterial blood pressure (Figure 2A). Tracheal contractions evoked by the autacoid were abolished by either atropine or bilateral vagotomy, confirming their parasympathetic and cholinergic nature (Figure 2C). Following atropine administration and precontraction of the trachealis with 1 µM PGF₂, intravenous histamine also induced parasympathetic relaxations of the trachealis (21 ± 7% reversal of the PGF₂ contraction). These tracheal relaxations were absent in vagotomized animals and in animals pretreated with trimethaphan (Figure 2D). Intravenous injections of vehicle (saline) failed to evoke cholinergic or NANC reflexes.

View larger version (16K): [in this window] [in a new window]   View larger version (11K): [in this window] [in a new window]   Figure 2.   Reflex-mediated, parasympathetic nerve-dependent contractions and relaxations of guinea pig trachealis in situ evoked by intravenous histamine. (A) Under control conditions, intravenous histamine evoked rapidly developing but short-lived increases in cholinergic tone in the trachea (TT), responses that coincided with the effects of histamine on pulmonary insufflation pressure (PT). Atropine abolished these histamine-mediated cholinergic reflexes in the trachea (the histamine H1 receptor antagonist pyrilamine was always present in the tracheal perfusate to prevent any direct effects of histamine on the trachealis). (B) Following atropine administration and precontraction of the trachealis with 1 µM PGF2, intravenous histamine also evoked reflex tracheal relaxations. (C ) Mean ± SEM effects of histamine on cholinergic tone. (D) Mean ± SEM effects of intravenous histamine on the PGF2- induced contractions. Each bar is the mean ± SEM of three to eight experiments.

Reflex Alterations in Tracheal Tension Evoked from the Larynx

Application of capsaicin to the laryngeal mucosa evoked reflex parasympathetic-cholinergic contractions of the trachealis. These increases in tracheal cholinergic tone evoked by laryngeal capsaicin were characteristically preceded by transient falls in tension (33 ± 9% reduction in baseline cholinergic tone) and occurred independent of any marked changes in PT or arterial blood pressure (Figure 3A). The contractions evoked by laryngeal capsaicin were slow in onset and were sustained for the duration of all experiments (up to 60 minutes). At the peak of the contraction, cholinergic tone in the trachea was elevated by 50-100% in most animals studied (n = 6). Tracheal instillation of atropine reversed the increase in tracheal tone (Figure 3A), while bilateral vagotomy or trimethaphan pretreatment prevented completely the effects of laryngeal capsaicin (Figure 3C), confirming the reflex and parasympathetic nature of the responses. As reported previously (19, 20), comparable contractile reflexes could be evoked by brief (1-2 minute) laryngeal application of either 3 µM bradykinin or acidic (0.1 N) saline. In contrast, 30 µM histamine applied to the laryngeal mucosa (administered when pyrilamine was left out of the tracheal/laryngeal perfusate) in the same manner failed to reflexively alter tracheal tension (n = 4).

View larger version (11K): [in this window] [in a new window]   View larger version (16K): [in this window] [in a new window]   View larger version (11K): [in this window] [in a new window]   Figure 3.   Reflex-mediated alterations in parasympathetic- cholinergic and -noncholinergic nerve-mediated contractions and relaxations of guinea pig trachealis in situ following brief (1-2 minute) application of capsaicin to the laryngeal mucosa. (A) Under control conditions, brief application of capsaicin to the laryngeal mucosa evoked slowly developing but long-lasting (up to 60 minutes) increases in tracheal cholinergic tone (TT). These increases in cholinergic tone were typically preceeded by rapidly developing but short-lived decreases in cholinergic tone. Local application of neurokinin receptor antagonists to the trachea and larynx had no effect on these laryngeal capsaicin-induced reflexes (unpublished observations; n = 4). The contractile reflexes were, however, reversed completely by adding atropine to the tracheal perfusate and prevented by either vagotomy or trimethaphan (see [C]). (B) Following atropine administration and precontraction of the trachealis with histamine, laryngeal capsaicin initiated tracheal relaxations that were abolished by vagotomy or trimethaphan. (C ) Mean peak effects of laryngeal capsaicin on baseline cholinergic tone. (D) Mean relaxant effects induced by laryngeal capsaicin. Each bar is the mean ± SEM of three to eight experiments.

Following atropine pretreatment and contraction of the trachealis with 10 µM of histamine, laryngeal application of 1-µM capsaicin also evoked relaxant reflexes in the guinea pig trachea (18 ± 4% reversal of the histamine contraction, n = 8; Figure 3B). The tracheal relaxations evoked by laryngeal capsaicin were abolished by bilateral vagotomy or trimethaphan pretreatment (Figure 3D). Laryngeal application of vehicle (Krebs buffer) failed to alter baseline cholinergic or NANC tone.

Reflex Alterations in Tracheal Tension Evoked by Nebulized Capsaicin and Histamine

We have previously shown that nebulized histamine has inconsistent effects on baseline cholinergic tone in the trachealis in situ despite evoking marked increases in PT (18). It is possible that inhaled histamine evokes reflex tracheal contractions and relaxations simultaneously, thus resulting in the inconsistent effects on cholinergic tone observed previously. Evidence against this hypothesis was gathered in the present study, however, as following atropine pretreatment and contraction of the trachealis with PGF₂, aerosolized histamine (10 µg/ml) was also ineffective at initiating reflex tracheal relaxations (5 ± 5% reversal of the PGF₂-induced contraction; n = 4). By contrast, aerosolized capsaicin (100 µM) evoked marked increases in cholinergic tone (45 ± 9% increase in cholinergic tone; n = 5) under control conditions, and following atropine pretreatment and contraction of the trachealis with histamine, reflex-mediated relaxations of the trachea (27 ± 7% reversal of the histamine contraction; n = 6). Either atropine pretreatment or vagotomy abolished the parasympathetic-cholinergic reflexes evoked by aerosolized capsaicin (n = 3-4). Bilateral vagotomy abolished the aerosol capsaicin-mediated parasympathetic relaxant responses (n = 3). Nebulized vehicle (saline) was without effect of parasympathetic nerve activity.

Reflexes Evoked by Increased Positive-End Expiratory Pressure or by Hypoxia

Transient increases in positive-end expiratory pressure (PEEP) produced decreases in baseline cholinergic tone (16 ± 3% reduction in baseline cholinergic tone, n = 4). The effects of increased PEEP on baseline cholinergic tone were rapid in onset, and persisted for the duration of the PEEP maneuver. Conversely, brief bouts of hypoxia (1-2 minutes of ventilation with 100% nitrogen gas) produced gradual increases in tracheal tone. Tracheal contractions evoked by hypoxia (72 ± 14% increase in cholinergic tone) were absent in vagotomized animals or animals pretreated with atropine. Interestingly, neither PEEP nor hypoxia evoked any apparent alterations in noncholinergic parasympathetic nerve activity in the trachea (Figure 4; Table 1).

View larger version (15K): [in this window] [in a new window]   View larger version (14K): [in this window] [in a new window]   Figure 4.   Representative traces of the effects of brief bouts of hypoxia on parasympathetic tone in the guinea pig trachea in situ. (A) Hypoxia (produced by ventilation with 100% nitrogen [N2] for 1-2 minutes) initiated slowly developing increases in tracheal cholinergic tone (TT), an effect that was prevented entirely by atropine pretreatment or vagotomy (see Table 1). Following atropine pretreatment and precontraction of the trachealis with histamine, hypoxia failed to produce an appreciable change in tone. Traces are representative of four to seven experiments. See text for further details.

Reflex Alterations in Tracheal Tension Evoked by Stimulation of Extrapulmonary Afferent Nerves

In an attempt to simulate the mammalian diving response, cold (4° C) saline was applied to the nose and nasal passages of anesthetized guinea pigs. This maneuver evoked a 48 ± 12% increase in tracheal cholinergic tone, an effect that was often accompanied by bradycardia and a fall in mean arterial blood pressure (8-15 mm Hg; n = 7). Tracheal contractions evoked during the diving response were prevented entirely by atropine (n = 6) or prior vagotomy (n = 3) and were not mimicked by replacing the cold saline with room temperature saline (n = 3). Activation of the diving reflex induced modest relaxations of the histamine precontracted trachea (5 ± 1% reversal of the histamine contraction, n = 5), an effect that was absent in vagotomized animals (Table 1).

In contrast to results of a previous study performed in dogs (24), unilateral stimulation (12-25 V, 1-millisecond pulse duration 5-10 Hz, 1-2 minutes) of a phrenic nerve failed to measurably alter autonomic tone in the guinea pig trachea (n = 3). On the other hand, bilateral stimulation of the vagus nerves caudal to the bronchi (adjacent to the esophagus, at the level of the diaphragm) consistently evoked contractions of the trachea (55 ± 19% increase in cholinergic tone, n = 5; Table 1, Figure 5]). These contractions were blocked entirely by atropine pretreatment, confirming their cholinergic nature. Following atropine pretreatment and precontraction of the trachealis with histamine, stimulating the subbronchial vagi also evoked noncholinergic relaxations of the trachealis (24 ± 4% reversal of the histamine contraction; n = 4), an effect that was prevented by sectioning the vagus nerves adjacent to the trachea.

View larger version (18K): [in this window] [in a new window]   View larger version (17K): [in this window] [in a new window]   Figure 5.   Effect of stimulating afferent nerves innervating the esophagus and other abdominal viscera on parasympathetic tone of the guinea pig trachealis in situ. (A) Under control conditions, bilateral stimulation of the vagus nerves caudal to the bronchi and adjacent to the diaphragm evoked marked and rapidly developing increases in cholinergic tone in the trachealis (TT), an effect that was prevented by atropine or by bilateral extrathoracic vagotomy. (B) Following atropine pretreatment and precontraction of the trachealis with histamine, electrical stimulation of the vagi caudal to the airways evoked frequency-dependent relaxations of the trachealis. These relaxations were prevented by prior extrathoracic vagotomy (see Table 1). Traces are representative of four experiments.

Kinetics of Contractile and Relaxant Reflexes in the Guinea Pig Airways

Even though contractile and relaxant reflexes were initiated in an identical manner, the time course of reflex-mediated contractions and relaxations evoked by histamine and capsaicin differed substantially. Whereas reflex-mediated cholinergic contractions evoked by intravenous histamine, inhaled capsaicin and hypoxia all reached half-maximal and maximal levels within 1-2 minutes postchallenge, relaxant responses evoked by these stimuli were much slower developing, requiring 2-4-fold more time to reach half maximal and maximal levels (Figures 2 and 6 and Table 2). This pattern was reversed entirely, however, when reflexes were initiated from the larynx by brief (1-2 minute) laryngeal mucosal application of capsaicin. Thus, as with intraveneous histamine and inhaled capsaicin, laryngeal capsaicin-induced relaxant reflexes developed and peaked within 1-2 minutes postchallenge, whereas reflex-mediated increases in cholinergic tone evoked by laryngeal capsaicin were exceedingly slow in developing, requiring up to 28 minutes to reach peak levels (Figures 3 and 6 and Table 2).

View larger version (27K): [in this window] [in a new window]   Figure 6.   Time course for reflex-mediated contractions (closed circles) and relaxations (open circles) evoked by (A) intravenous histamine (10 µg/ kg), (B) inhaled capsaicin (100 µM), (C ) hypoxia (1-2 minute ventilation of 100% nitrogen), and (D) laryngeal mucosal application of capsaicin. Results are presented as a percentage of the maximum response attained following stimulation by the various methods. Contractile reflexes evoked by the various stimuli were comparable in magnitude (see Figures 2 and 3; also see text). Relaxant responses evoked by capsaicin (inhaled, laryngeal) or intravenous histamine were comparable in magnitude, while hypoxia apparently failed to evoke any substantive change in noncholinergic parasympathetic nerve activity in the airways. Each point is the mean ± SEM of four to eight experiments.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In many species, including humans, the parasympathetic nervous system provides both contractile and relaxant innervation to airway smooth muscle (1, 3, 25). Data outlined in the present study and elsewhere indicate that cholinergic and noncholinergic parasympathetic nerves projecting to the airways are regulated by several afferent inputs to the central nervous system. The data presented also indicate that parasympathetic contractile and relaxant pathways innervating guinea pig airways may be subject to differential reflex regulation. The findings further highlight the distinctions between the cholinergic and noncholinergic parasympathetic pathways innervating the airways and provide important insights into autonomic regulation of bronchomotor tone.

Evidence for Anatomically and Physiologically Distinct Vagal Contractile and Relaxant Pathways Innervating Airway Smooth Muscle

Previous studies in guinea pigs indicate that airway cholinergic and noncholinergic parasympathetic nerves represent distinct vagal pathways rather than a single neural pathway mediating both contractile and relaxant responses (4-8). Immunohistochemical analyses of autonomic ganglia intrinsic to the guinea pig airway fail to provide evidence for corelease of acetylcholine and putative relaxant neurotransmitters (5-8, 26-29). Rather, studies reveal that vasoactive intestinal peptide/nitric oxide synthase-containing parasympathetic nerves innervating the guinea pig airways originate in the nearby esophagus (5, 7, 8). Indeed, disrupting the tissue between the esophagus and trachea abolishes vagally mediated relaxant responses in the trachea, while preserving cholinergic contractions (4). Electrophysiologic studies in cats and guinea pigs provide further evidence for distinct contractile and relaxant parasympathetic pathways (4, 10). In both species, myelinated preganglionic nerve fibers appear to mediate cholinergic contractions, while unmyelinated preganglionic nerve fibers regulate noncholinergic relaxations. Circumstantial evidence as well as direct immunohistochemical evidence suggests a comparable arrangement of the parasympathetic innervation in ferret and human airways (13-16).

In the present study we observed that stimuli selective for airway C-fibers (capsaicin) and airway rapidly adapting receptors (histamine) were equally effective at reflexively activating cholinergic and noncholinergic pathways. Bradykinin, another selective C-fiber stimulant, also initiates both cholinergic and noncholinergic parasympathetic reflexes when delivered by aerosol, intravenously, or applied onto the laryngeal mucosa (18, 20). Cholinergic reflexes comparable in magnitude to those evoked by intravenous histamine or inhaled capsaicin could also be evoked during brief bouts of hypoxia or during the diving response evoked by cooling the nose. Conversely, slowly adapting receptor activation (increased PEEP) reduced baseline cholinergic tone. In sharp contrast, however, we observed that presumed chemoreceptor activation (hypoxia), PEEP and the diving reflex all apparently failed to produce appreciable changes (increases or decreases) in noncholinergic parasympathetic nerve activity. Comparable results have been reported in studies using cats (11, 30). We thus conclude that reflexes differentially regulate cholinergic and noncholinergic parasympathetic nerves in the airways.

Further evidence for differential regulation of parasympathetic contractile and relaxant pathways comes from studying the time course of the reflexes. The time to onset, time to peak effects, and thus rate effects of the contractile reflexes evoked by inhaled capsaicin and intravenous histamine were consistently shorter and faster than that for the relaxant reflexes evoked by these stimuli. Similar results were obtained previously using bradykinin (intravenous, inhaled [20]). The stimuli (inhaled capsaicin, intravenous histamine) were delivered in an identical manner when contractions and relaxations were studied, and thus the afferent activity was likely to be identical in these parallel experiments. It is thus likely that the different kinetics for the contractile and relaxant responses observed are not due to differences in the stimulus delivered. Rather, the kinetic differences likely reside in the efferent limbs of the responses. The disparities in half-lives and time to peak for the contractile and relaxant reflexes can be attributed at least in part to the different kinetics of the responses at the level of the smooth muscle. Vagal stimulation experiments clearly illustrate the more rapid onset and attainment of equilibrium of the contractile response of the trachealis than comparably evoked relaxant responses ([4, 31] present study). When contractile and relaxant reflexes were evoked from the larynx, however, completely opposite kinetics were observed. Relaxations evoked by laryngeal capsaicin occurred at a rate similar to the relaxant reflexes evoked by inhaled capsaicin or intravenous histamine. Also, like the relaxant reflexes evoked by inhaled capsaicin or intravenous histamine, the relaxations evoked by laryngeal capsaicin slowly but persistently recovered over the 5-10 minutes following the brief laryngeal capsaicin challenge. By contrast, the contractile reflexes evoked by laryngeal capsaicin were exceedingly slow in onset and long lasting. Although the mechanisms accounting for these kinetics are not clear, the data provide further evidence that the cholinergic and noncholinergic parasympathetic nerves are differentially regulated by reflexes.

Regulation of Airway Cholinergic and Noncholinergic Parasympathetic Nerve Activity by Activation of Extrapulmonary Afferent Nerves

Conditions such as gastroesophageal reflux disease, upper respiratory tract infection, and rhinitis can precipitate pulmonary responses and symptoms such as bronchospasm, dyspnea, and cough (32-34). These pulmonary symptoms often occur independent of any inflammation of the lower airways or any direct action of the refluxate, infectious pathogens, or postnasal drip with the structures innervated by airway afferent nerves. Nevertheless, in gastroesophageal reflux disease and upper respiratory tract infection (and probably rhinitis), the accompanying bronchospasm and airways hyperresponsiveness are mediated by exaggerated parasympathetic-cholinergic reflexes (32-34). In the present study we confirmed the ability of extrapulmonary afferent nerve (upper airway, esophageal, gastrointestinal) activation to initiate parasympathetic-cholinergic reflexes. We extended this observation by showing that activation of extrapulmonary afferent nerves also initiates parasympathetic-noncholinergic reflexes in the airways. What, if any, role these reflexes evoked by extrapulmonary stimulation play in homeostasis is not readily apparent. Perhaps the pulmonary reflexes initiated by activation of esophageal and/or upper airway afferent nerves serve to coordinate defensive reflexes such as sneezing or coughing associated with aspiration or postnasal drip. Alternatively, the reflexes might occur only as a consequence of the imprecision with which the autonomic nervous system regulates the viscera. Indeed, discrete regions of brain stem nuclei are known to regulate multiple visceral organ functions (35). This anatomical overlap might account for the cluster of autonomic responses (alterations in airway caliber, changes in heart rate and/or blood pressure, alterations in gastrointestinal motility) associated with activation of specific afferent nerve subtypes.

It is noteworthy that many stimuli used to evoke profound alterations in parasympathetic nerve activity as measured in the trachea had little or no measurable effects on PT. Although there may be differences in the expression of muscarinic receptors throughout the tracheobronchial tree (36), it is unlikely that parasympathetic regulation of tracheal function differs substantially from parasympathetic regulation of the lower airways (3, 17-19). Rather, we interpret this observation as further evidence of the bluntness of the PT measurement as a tool for studying autonomic regulation of airway tone. There are many other disadvantages to using whole-lung mechanics for studying airway autonomic nerves. It is essentially impossible, for example, to effectively isolate stimuli (e.g., inhaled irritants, alterations in tidal volume, end-expiratory pressure, respiratory rate, or gas mixture) from the responding airways using whole lung mechanics. Also problematic are the potential modulatory effects of stimuli on synaptic transmission in airway ganglia or on postganglionic nerve endings that might produce responses that appear to be reflexive when they are purely peripherally mediated (3, 37). These limitations further highlights the many strengths of our model where stimuli (e.g., laryngeal or inhaled capsaicin, intravenous histamine) can be physically or pharmacologically (by intratracheal administration of appropriate receptor antagonists) isolated from the responding end organ (the extrathoracic trachea) and effector efferent nerves.

Potential Relevance of Differential Reflex Regulation of Airway Parasympathetic Nerve Activity in Health and Disease

Airways hyperreactivity is a common presenting symptom of asthma, upper respiratory tract infection, rhinitis, and gastroesophageal reflux disease. The airways hyperreactivity associated with these conditions can often be reversed by anticholinergic agents such as atropine or ipratropium bromide, suggesting a prominent role for the autonomic nervous system in airways dysfunction (32-34). Altered cholinergic nerve activity may not, however, be the sole neural mechanism underlying airways hyperreactivity. Dysfunction of noncholinergic parasympathetic nerves may also be a contributing factor (1- 3, 38). Given that the cholinergic and noncholinergic parasympathetic pathways innervating the airways seem to be independently regulated, dysregulation of either pathway could presumably precipitate airways hyperresponsiveness.

We speculate that differential regulation of cholinergic and noncholinergic parasympathetic nerves of the airways reflects their relative importance in optimizing defensive and vegetative reflexes. The almost instantaneous kinetics of the contractions and relaxations produced by increasing or decreasing cholinergic nerve activity permits breath to breath alterations in airway caliber that may optimize the work of breathing by matching demands for gas exchange with the benefits of limiting dead space. When demands are low during rest and tidal breathing, for example, minimizing dead space by decreasing airway caliber optimizes gas exchange by reducing the work of breathing. Conversely, increased demand during exercise can be met by increasing minute ventilation (by increasing tidal volume and respiratory rate) and decreasing airways resistance following withdrawal of baseline cholinergic tone. Stimulation of skeletal muscle afferent nerves to mimic exercise initiates withdrawal of baseline cholinergic tone (23, 39). The cholinergic nerves may also play an important role in defensive reflexes. Rapid increases in cholinergic tone during coughing, for example, might facilitate clearance of inhaled irritants, mucus, or particulate matter by increasing expiratory velocity. The exceedingly slow kinetics of noncholinergic relaxant responses renders this system less useful in vegetative reflexes. By contrast, however, with defensive reflexes such as coughing, the relaxant responses might serve to normalize airway tone once the irritating or obstructive substance is cleared.

Finally, the results of the present study highlight the importance of careful experimental design when studying the role of noncholinergic parasympathetic nerves in regulating airways reactivity. Measuring only the peak excitatory effects of stimuli on tracheal tone before and after inhibition of the noncholinergic nerves would likely fail to reveal a modulatory role of the noncholinergic nerves due to the temporal dissociation between the contractile and relaxant reflex effects. Indeed, we have shown that although inhibition of NANC nerve-mediated effects using the nitric oxide synthase inhibitor L-NNA or the soluble guanylate cyclase inhibitor ODQ increases baseline cholinergic tone, these interventions have no effect on bradykinin-mediated contractile reflexes (20). Obviously, PT measurements or other measurements of whole lung mechanics would have similar and additional problems associated with the direct effects of stimuli on lung mechanics. These issues should be considered in future studies aimed at defining the role of airway parasympathetic nerves in the pathogenesis of airways hyperresponsiveness.